picoclaw/pkg/agent/loop.go

1704 lines
51 KiB
Go
Raw Normal View History

2026-02-04 11:06:13 +00:00
// PicoClaw - Ultra-lightweight personal AI agent
// Inspired by and based on nanobot: https://github.com/HKUDS/nanobot
// License: MIT
//
// Copyright (c) 2026 PicoClaw contributors
package agent
import (
"context"
"encoding/json"
"errors"
2026-02-04 11:06:13 +00:00
"fmt"
"path/filepath"
2026-03-04 10:21:59 +00:00
"regexp"
"strings"
"sync"
"sync/atomic"
"time"
"unicode/utf8"
2026-02-04 11:06:13 +00:00
"github.com/sipeed/picoclaw/pkg/bus"
"github.com/sipeed/picoclaw/pkg/channels"
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
"github.com/sipeed/picoclaw/pkg/commands"
2026-02-04 11:06:13 +00:00
"github.com/sipeed/picoclaw/pkg/config"
"github.com/sipeed/picoclaw/pkg/constants"
"github.com/sipeed/picoclaw/pkg/logger"
"github.com/sipeed/picoclaw/pkg/mcp"
"github.com/sipeed/picoclaw/pkg/media"
2026-02-04 11:06:13 +00:00
"github.com/sipeed/picoclaw/pkg/providers"
"github.com/sipeed/picoclaw/pkg/routing"
"github.com/sipeed/picoclaw/pkg/skills"
"github.com/sipeed/picoclaw/pkg/state"
2026-02-04 11:06:13 +00:00
"github.com/sipeed/picoclaw/pkg/tools"
"github.com/sipeed/picoclaw/pkg/utils"
2026-03-01 08:31:04 +00:00
"github.com/sipeed/picoclaw/pkg/voice"
2026-02-04 11:06:13 +00:00
)
type AgentLoop struct {
bus *bus.MessageBus
cfg *config.Config
registry *AgentRegistry
state *state.Manager
running atomic.Bool
summarizing sync.Map
fallback *providers.FallbackChain
channelManager *channels.Manager
mediaStore media.MediaStore
2026-03-01 08:31:04 +00:00
transcriber voice.Transcriber
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
cmdRegistry *commands.Registry
2026-02-04 11:06:13 +00:00
}
// processOptions configures how a message is processed
type processOptions struct {
SessionKey string // Session identifier for history/context
Channel string // Target channel for tool execution
ChatID string // Target chat ID for tool execution
UserMessage string // User message content (may include prefix)
Media []string // media:// refs from inbound message
DefaultResponse string // Response when LLM returns empty
EnableSummary bool // Whether to trigger summarization
SendResponse bool // Whether to send response via bus
NoHistory bool // If true, don't load session history (for heartbeat)
2026-02-04 11:06:13 +00:00
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
const (
defaultResponse = "I've completed processing but have no response to give. Increase `max_tool_iterations` in config.json."
sessionKeyAgentPrefix = "agent:"
metadataKeyAccountID = "account_id"
metadataKeyGuildID = "guild_id"
metadataKeyTeamID = "team_id"
metadataKeyParentPeerKind = "parent_peer_kind"
metadataKeyParentPeerID = "parent_peer_id"
)
func NewAgentLoop(
cfg *config.Config,
msgBus *bus.MessageBus,
provider providers.LLMProvider,
) *AgentLoop {
registry := NewAgentRegistry(cfg, provider)
// Register shared tools to all agents
registerSharedTools(cfg, msgBus, registry, provider)
// Set up shared fallback chain
cooldown := providers.NewCooldownTracker()
fallbackChain := providers.NewFallbackChain(cooldown)
// Create state manager using default agent's workspace for channel recording
defaultAgent := registry.GetDefaultAgent()
var stateManager *state.Manager
if defaultAgent != nil {
stateManager = state.NewManager(defaultAgent.Workspace)
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
al := &AgentLoop{
bus: msgBus,
cfg: cfg,
registry: registry,
state: stateManager,
summarizing: sync.Map{},
fallback: fallbackChain,
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
cmdRegistry: commands.NewRegistry(commands.BuiltinDefinitions()),
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
return al
}
2026-02-04 11:06:13 +00:00
// registerSharedTools registers tools that are shared across all agents (web, message, spawn).
2026-02-20 18:03:11 +00:00
func registerSharedTools(
cfg *config.Config,
msgBus *bus.MessageBus,
registry *AgentRegistry,
provider providers.LLMProvider,
) {
for _, agentID := range registry.ListAgentIDs() {
agent, ok := registry.GetAgent(agentID)
if !ok {
continue
}
// Web tools
if cfg.Tools.IsToolEnabled("web") {
searchTool, err := tools.NewWebSearchTool(tools.WebSearchToolOptions{
BraveAPIKey: cfg.Tools.Web.Brave.APIKey,
BraveMaxResults: cfg.Tools.Web.Brave.MaxResults,
BraveEnabled: cfg.Tools.Web.Brave.Enabled,
TavilyAPIKey: cfg.Tools.Web.Tavily.APIKey,
TavilyBaseURL: cfg.Tools.Web.Tavily.BaseURL,
TavilyMaxResults: cfg.Tools.Web.Tavily.MaxResults,
TavilyEnabled: cfg.Tools.Web.Tavily.Enabled,
DuckDuckGoMaxResults: cfg.Tools.Web.DuckDuckGo.MaxResults,
DuckDuckGoEnabled: cfg.Tools.Web.DuckDuckGo.Enabled,
PerplexityAPIKey: cfg.Tools.Web.Perplexity.APIKey,
PerplexityMaxResults: cfg.Tools.Web.Perplexity.MaxResults,
PerplexityEnabled: cfg.Tools.Web.Perplexity.Enabled,
SearXNGBaseURL: cfg.Tools.Web.SearXNG.BaseURL,
SearXNGMaxResults: cfg.Tools.Web.SearXNG.MaxResults,
SearXNGEnabled: cfg.Tools.Web.SearXNG.Enabled,
GLMSearchAPIKey: cfg.Tools.Web.GLMSearch.APIKey,
GLMSearchBaseURL: cfg.Tools.Web.GLMSearch.BaseURL,
GLMSearchEngine: cfg.Tools.Web.GLMSearch.SearchEngine,
GLMSearchMaxResults: cfg.Tools.Web.GLMSearch.MaxResults,
GLMSearchEnabled: cfg.Tools.Web.GLMSearch.Enabled,
Proxy: cfg.Tools.Web.Proxy,
})
if err != nil {
logger.ErrorCF("agent", "Failed to create web search tool", map[string]any{"error": err.Error()})
} else if searchTool != nil {
agent.Tools.Register(searchTool)
}
}
if cfg.Tools.IsToolEnabled("web_fetch") {
fetchTool, err := tools.NewWebFetchToolWithProxy(50000, cfg.Tools.Web.Proxy, cfg.Tools.Web.FetchLimitBytes)
if err != nil {
logger.ErrorCF("agent", "Failed to create web fetch tool", map[string]any{"error": err.Error()})
} else {
agent.Tools.Register(fetchTool)
}
fix(tools): close resp.Body on retry cancel and cache http.Client instances (#940) * fix(tools): close resp.Body on retry cancel and cache http.Client instances Fix resp.Body leak in DoRequestWithRetry where req.Body (request) was incorrectly closed instead of resp.Body (response) on context cancel. Cache http.Client on web search/fetch provider structs and channel adapters (WeCom, LINE) to avoid per-call allocation overhead. * fix(channels): preserve original http client timeouts for LINE and WeCom Split LINE single 60s client into infoClient (10s) for bot info lookups and apiClient (30s) for messaging API calls. Lower WeCom cached client base timeout from 60s to 30s (matching uploadMedia), and ensure it is always >= the configured ReplyTimeout so the per-request context deadline remains the effective limit. * refactor(tools): extract timeout consts and deduplicate WebFetchTool constructors Address PR review feedback from xiaket: - Define searchTimeout, perplexityTimeout, fetchTimeout, defaultMaxChars, and maxRedirects as package-level consts instead of magic numbers. - Remove misleading "No proxy" comment in NewWebFetchTool. - Deduplicate NewWebFetchTool by delegating to NewWebFetchToolWithProxy. * test(utils): add context cancellation test for DoRequestWithRetry Verify that resp.Body is properly closed when the context is canceled during retry sleep, covering the C8 resp.Body leak fix. * fix(utils): close resp in test to satisfy bodyclose linter * fix(utils): eliminate flakiness in context cancellation retry test Synchronize cancellation using an onRoundTrip callback from the transport wrapper instead of a timing-based context timeout. This ensures the first client.Do completes before cancel fires, so cancellation always hits during sleepWithCtx.
2026-03-01 05:55:46 +00:00
}
// Hardware tools (I2C, SPI) - Linux only, returns error on other platforms
if cfg.Tools.IsToolEnabled("i2c") {
agent.Tools.Register(tools.NewI2CTool())
}
if cfg.Tools.IsToolEnabled("spi") {
agent.Tools.Register(tools.NewSPITool())
}
// Message tool
if cfg.Tools.IsToolEnabled("message") {
messageTool := tools.NewMessageTool()
messageTool.SetSendCallback(func(channel, chatID, content string) error {
pubCtx, pubCancel := context.WithTimeout(context.Background(), 5*time.Second)
defer pubCancel()
return msgBus.PublishOutbound(pubCtx, bus.OutboundMessage{
Channel: channel,
ChatID: chatID,
Content: content,
})
})
agent.Tools.Register(messageTool)
}
feat(feishu,tools): add outbound media delivery via send_file tool (#1156) * feat(feishu): implement SendMedia and add send_file tool Add outbound media support for the Feishu channel so the agent can send images and files to users via the MediaStore pipeline. Feishu channel: - SendMedia dispatches media parts as image or file uploads - sendImage uploads via Image.Create then sends image message - sendFile uploads via File.Create then sends file message - feishuFileType maps extensions to Feishu file_type values send_file tool: - New tool lets the LLM send a local file to the current chat - Validates path, registers file in MediaStore, returns media ref - Agent loop wires tool registration, MediaStore propagation, and context updates Tested on Radxa Cubie A7A (arm64) with Feishu websocket channel. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): publish outbound media regardless of SendResponse flag The SendResponse flag controls whether the agent loop publishes the final text response (callers that publish it themselves set this to false). However, the media publish path was also gated behind this flag, which meant tool-produced media was silently dropped for normal channel messages. Media should be published immediately when a tool returns media refs, independent of how the text response is delivered. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(tools): use magic-bytes MIME detection and add file size limit to send_file - Replace hardcoded extension-to-MIME map with h2non/filetype (magic bytes) + mime.TypeByExtension fallback, consistent with the vision pipeline in resolveMediaRefs - Add configurable max file size check (defaults to config.DefaultMaxMediaSize, 20 MB) to prevent oversized uploads - Add tests for magic-bytes detection, extension fallback, size limit, and default max size Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): add ForEachTool to AgentRegistry for cross-agent tool lookup Extract the pattern of iterating agents to find a named tool into AgentRegistry.ForEachTool, simplifying SetMediaStore propagation. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent,tools): adapt send_file to ctx-based channel injection after upstream refactor Replace ContextualTool interface (removed upstream) with direct ctx reading in SendFileTool.Execute, using ToolChannel/ToolChatID helpers. Remove updateToolContexts which is no longer needed since ExecuteWithContext already injects channel/chatID into ctx for all tools. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(tools): support toggling send_file tool via config Add SendFileConfig with Enabled field to ToolsConfig, defaulting to true. Wrap send_file tool registration in loop.go with the config check, consistent with the pattern used by other tools. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 11:42:52 +00:00
// Send file tool (outbound media via MediaStore — store injected later by SetMediaStore)
if cfg.Tools.IsToolEnabled("send_file") {
sendFileTool := tools.NewSendFileTool(
agent.Workspace,
cfg.Agents.Defaults.RestrictToWorkspace,
cfg.Agents.Defaults.GetMaxMediaSize(),
nil,
)
agent.Tools.Register(sendFileTool)
}
// Skill discovery and installation tools
skills_enabled := cfg.Tools.IsToolEnabled("skills")
find_skills_enable := cfg.Tools.IsToolEnabled("find_skills")
install_skills_enable := cfg.Tools.IsToolEnabled("install_skill")
if skills_enabled && (find_skills_enable || install_skills_enable) {
registryMgr := skills.NewRegistryManagerFromConfig(skills.RegistryConfig{
MaxConcurrentSearches: cfg.Tools.Skills.MaxConcurrentSearches,
ClawHub: skills.ClawHubConfig(cfg.Tools.Skills.Registries.ClawHub),
})
if find_skills_enable {
searchCache := skills.NewSearchCache(
cfg.Tools.Skills.SearchCache.MaxSize,
time.Duration(cfg.Tools.Skills.SearchCache.TTLSeconds)*time.Second,
)
agent.Tools.Register(tools.NewFindSkillsTool(registryMgr, searchCache))
}
if install_skills_enable {
agent.Tools.Register(tools.NewInstallSkillTool(registryMgr, agent.Workspace))
}
}
// Spawn tool with allowlist checker
if cfg.Tools.IsToolEnabled("spawn") {
if cfg.Tools.IsToolEnabled("subagent") {
subagentManager := tools.NewSubagentManager(provider, agent.Model, agent.Workspace)
subagentManager.SetLLMOptions(agent.MaxTokens, agent.Temperature)
spawnTool := tools.NewSpawnTool(subagentManager)
currentAgentID := agentID
spawnTool.SetAllowlistChecker(func(targetAgentID string) bool {
return registry.CanSpawnSubagent(currentAgentID, targetAgentID)
})
agent.Tools.Register(spawnTool)
} else {
logger.WarnCF("agent", "spawn tool requires subagent to be enabled", nil)
}
}
2026-02-04 11:06:13 +00:00
}
}
func (al *AgentLoop) Run(ctx context.Context) error {
al.running.Store(true)
2026-02-04 11:06:13 +00:00
// Initialize MCP servers for all agents
if al.cfg.Tools.IsToolEnabled("mcp") {
mcpManager := mcp.NewManager()
// Ensure MCP connections are cleaned up on exit, regardless of initialization success
// This fixes resource leak when LoadFromMCPConfig partially succeeds then fails
defer func() {
if err := mcpManager.Close(); err != nil {
logger.ErrorCF("agent", "Failed to close MCP manager",
map[string]any{
"error": err.Error(),
})
}
}()
defaultAgent := al.registry.GetDefaultAgent()
var workspacePath string
if defaultAgent != nil && defaultAgent.Workspace != "" {
workspacePath = defaultAgent.Workspace
} else {
workspacePath = al.cfg.WorkspacePath()
}
if err := mcpManager.LoadFromMCPConfig(ctx, al.cfg.Tools.MCP, workspacePath); err != nil {
logger.WarnCF("agent", "Failed to load MCP servers, MCP tools will not be available",
map[string]any{
"error": err.Error(),
})
} else {
// Register MCP tools for all agents
servers := mcpManager.GetServers()
uniqueTools := 0
totalRegistrations := 0
agentIDs := al.registry.ListAgentIDs()
agentCount := len(agentIDs)
for serverName, conn := range servers {
uniqueTools += len(conn.Tools)
for _, tool := range conn.Tools {
for _, agentID := range agentIDs {
agent, ok := al.registry.GetAgent(agentID)
if !ok {
continue
}
mcpTool := tools.NewMCPTool(mcpManager, serverName, tool)
agent.Tools.Register(mcpTool)
totalRegistrations++
logger.DebugCF("agent", "Registered MCP tool",
map[string]any{
"agent_id": agentID,
"server": serverName,
"tool": tool.Name,
"name": mcpTool.Name(),
})
}
}
}
logger.InfoCF("agent", "MCP tools registered successfully",
map[string]any{
"server_count": len(servers),
"unique_tools": uniqueTools,
"total_registrations": totalRegistrations,
"agent_count": agentCount,
})
}
}
for al.running.Load() {
2026-02-04 11:06:13 +00:00
select {
case <-ctx.Done():
return nil
default:
msg, ok := al.bus.ConsumeInbound(ctx)
if !ok {
continue
}
// Process message
func() {
// TODO: Re-enable media cleanup after inbound media is properly consumed by the agent.
// Currently disabled because files are deleted before the LLM can access their content.
// defer func() {
// if al.mediaStore != nil && msg.MediaScope != "" {
// if releaseErr := al.mediaStore.ReleaseAll(msg.MediaScope); releaseErr != nil {
// logger.WarnCF("agent", "Failed to release media", map[string]any{
// "scope": msg.MediaScope,
// "error": releaseErr.Error(),
// })
// }
// }
// }()
response, err := al.processMessage(ctx, msg)
if err != nil {
response = fmt.Sprintf("Error processing message: %v", err)
}
if response != "" {
// Check if the message tool already sent a response during this round.
// If so, skip publishing to avoid duplicate messages to the user.
// Use default agent's tools to check (message tool is shared).
alreadySent := false
defaultAgent := al.registry.GetDefaultAgent()
if defaultAgent != nil {
if tool, ok := defaultAgent.Tools.Get("message"); ok {
if mt, ok := tool.(*tools.MessageTool); ok {
alreadySent = mt.HasSentInRound()
}
}
}
if !alreadySent {
al.bus.PublishOutbound(ctx, bus.OutboundMessage{
Channel: msg.Channel,
ChatID: msg.ChatID,
Content: response,
})
logger.InfoCF("agent", "Published outbound response",
map[string]any{
"channel": msg.Channel,
"chat_id": msg.ChatID,
"content_len": len(response),
})
} else {
logger.DebugCF(
"agent",
"Skipped outbound (message tool already sent)",
map[string]any{"channel": msg.Channel},
)
}
}
}()
2026-02-04 11:06:13 +00:00
}
}
return nil
}
func (al *AgentLoop) Stop() {
al.running.Store(false)
2026-02-04 11:06:13 +00:00
}
func (al *AgentLoop) RegisterTool(tool tools.Tool) {
for _, agentID := range al.registry.ListAgentIDs() {
if agent, ok := al.registry.GetAgent(agentID); ok {
agent.Tools.Register(tool)
}
}
2026-02-04 11:06:13 +00:00
}
func (al *AgentLoop) SetChannelManager(cm *channels.Manager) {
al.channelManager = cm
}
// SetMediaStore injects a MediaStore for media lifecycle management.
func (al *AgentLoop) SetMediaStore(s media.MediaStore) {
al.mediaStore = s
feat(feishu,tools): add outbound media delivery via send_file tool (#1156) * feat(feishu): implement SendMedia and add send_file tool Add outbound media support for the Feishu channel so the agent can send images and files to users via the MediaStore pipeline. Feishu channel: - SendMedia dispatches media parts as image or file uploads - sendImage uploads via Image.Create then sends image message - sendFile uploads via File.Create then sends file message - feishuFileType maps extensions to Feishu file_type values send_file tool: - New tool lets the LLM send a local file to the current chat - Validates path, registers file in MediaStore, returns media ref - Agent loop wires tool registration, MediaStore propagation, and context updates Tested on Radxa Cubie A7A (arm64) with Feishu websocket channel. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): publish outbound media regardless of SendResponse flag The SendResponse flag controls whether the agent loop publishes the final text response (callers that publish it themselves set this to false). However, the media publish path was also gated behind this flag, which meant tool-produced media was silently dropped for normal channel messages. Media should be published immediately when a tool returns media refs, independent of how the text response is delivered. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(tools): use magic-bytes MIME detection and add file size limit to send_file - Replace hardcoded extension-to-MIME map with h2non/filetype (magic bytes) + mime.TypeByExtension fallback, consistent with the vision pipeline in resolveMediaRefs - Add configurable max file size check (defaults to config.DefaultMaxMediaSize, 20 MB) to prevent oversized uploads - Add tests for magic-bytes detection, extension fallback, size limit, and default max size Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): add ForEachTool to AgentRegistry for cross-agent tool lookup Extract the pattern of iterating agents to find a named tool into AgentRegistry.ForEachTool, simplifying SetMediaStore propagation. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent,tools): adapt send_file to ctx-based channel injection after upstream refactor Replace ContextualTool interface (removed upstream) with direct ctx reading in SendFileTool.Execute, using ToolChannel/ToolChatID helpers. Remove updateToolContexts which is no longer needed since ExecuteWithContext already injects channel/chatID into ctx for all tools. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(tools): support toggling send_file tool via config Add SendFileConfig with Enabled field to ToolsConfig, defaulting to true. Wrap send_file tool registration in loop.go with the config check, consistent with the pattern used by other tools. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 11:42:52 +00:00
// Propagate store to send_file tools in all agents.
al.registry.ForEachTool("send_file", func(t tools.Tool) {
if sf, ok := t.(*tools.SendFileTool); ok {
sf.SetMediaStore(s)
}
})
}
2026-03-01 08:31:04 +00:00
// SetTranscriber injects a voice transcriber for agent-level audio transcription.
func (al *AgentLoop) SetTranscriber(t voice.Transcriber) {
al.transcriber = t
}
var audioAnnotationRe = regexp.MustCompile(`\[(voice|audio)(?::[^\]]*)?\]`)
// transcribeAudioInMessage resolves audio media refs, transcribes them, and
// replaces audio annotations in msg.Content with the transcribed text.
func (al *AgentLoop) transcribeAudioInMessage(ctx context.Context, msg bus.InboundMessage) bus.InboundMessage {
if al.transcriber == nil || al.mediaStore == nil || len(msg.Media) == 0 {
2026-03-01 08:31:04 +00:00
return msg
}
// Transcribe each audio media ref in order.
var transcriptions []string
for _, ref := range msg.Media {
path, meta, err := al.mediaStore.ResolveWithMeta(ref)
if err != nil {
logger.WarnCF("voice", "Failed to resolve media ref", map[string]any{"ref": ref, "error": err})
continue
}
if !utils.IsAudioFile(meta.Filename, meta.ContentType) {
continue
}
result, err := al.transcriber.Transcribe(ctx, path)
if err != nil {
logger.WarnCF("voice", "Transcription failed", map[string]any{"ref": ref, "error": err})
transcriptions = append(transcriptions, "")
continue
}
transcriptions = append(transcriptions, result.Text)
}
if len(transcriptions) == 0 {
return msg
}
// Replace audio annotations sequentially with transcriptions.
idx := 0
newContent := audioAnnotationRe.ReplaceAllStringFunc(msg.Content, func(match string) string {
if idx >= len(transcriptions) {
return match
}
text := transcriptions[idx]
idx++
return "[voice: " + text + "]"
})
// Append any remaining transcriptions not matched by an annotation.
for ; idx < len(transcriptions); idx++ {
newContent += "\n[voice: " + transcriptions[idx] + "]"
}
msg.Content = newContent
return msg
}
// inferMediaType determines the media type ("image", "audio", "video", "file")
// from a filename and MIME content type.
func inferMediaType(filename, contentType string) string {
ct := strings.ToLower(contentType)
fn := strings.ToLower(filename)
if strings.HasPrefix(ct, "image/") {
return "image"
}
if strings.HasPrefix(ct, "audio/") || ct == "application/ogg" {
return "audio"
}
if strings.HasPrefix(ct, "video/") {
return "video"
}
// Fallback: infer from extension
ext := filepath.Ext(fn)
switch ext {
case ".jpg", ".jpeg", ".png", ".gif", ".webp", ".bmp", ".svg":
return "image"
case ".mp3", ".wav", ".ogg", ".m4a", ".flac", ".aac", ".wma", ".opus":
return "audio"
case ".mp4", ".avi", ".mov", ".webm", ".mkv":
return "video"
}
return "file"
}
// RecordLastChannel records the last active channel for this workspace.
// This uses the atomic state save mechanism to prevent data loss on crash.
func (al *AgentLoop) RecordLastChannel(channel string) error {
if al.state == nil {
return nil
}
return al.state.SetLastChannel(channel)
}
// RecordLastChatID records the last active chat ID for this workspace.
// This uses the atomic state save mechanism to prevent data loss on crash.
func (al *AgentLoop) RecordLastChatID(chatID string) error {
if al.state == nil {
return nil
}
return al.state.SetLastChatID(chatID)
}
func (al *AgentLoop) ProcessDirect(
ctx context.Context,
content, sessionKey string,
) (string, error) {
return al.ProcessDirectWithChannel(ctx, content, sessionKey, "cli", "direct")
}
2026-02-20 18:03:11 +00:00
func (al *AgentLoop) ProcessDirectWithChannel(
ctx context.Context,
content, sessionKey, channel, chatID string,
) (string, error) {
2026-02-04 11:06:13 +00:00
msg := bus.InboundMessage{
Channel: channel,
SenderID: "cron",
ChatID: chatID,
2026-02-04 11:06:13 +00:00
Content: content,
SessionKey: sessionKey,
}
return al.processMessage(ctx, msg)
}
// ProcessHeartbeat processes a heartbeat request without session history.
// Each heartbeat is independent and doesn't accumulate context.
func (al *AgentLoop) ProcessHeartbeat(
ctx context.Context,
content, channel, chatID string,
) (string, error) {
agent := al.registry.GetDefaultAgent()
if agent == nil {
return "", fmt.Errorf("no default agent for heartbeat")
}
return al.runAgentLoop(ctx, agent, processOptions{
SessionKey: "heartbeat",
Channel: channel,
ChatID: chatID,
UserMessage: content,
DefaultResponse: defaultResponse,
EnableSummary: false,
SendResponse: false,
NoHistory: true, // Don't load session history for heartbeat
})
}
2026-02-04 11:06:13 +00:00
func (al *AgentLoop) processMessage(ctx context.Context, msg bus.InboundMessage) (string, error) {
// Add message preview to log (show full content for error messages)
var logContent string
if strings.Contains(msg.Content, "Error:") || strings.Contains(msg.Content, "error") {
logContent = msg.Content // Full content for errors
} else {
logContent = utils.Truncate(msg.Content, 80)
}
logger.InfoCF(
"agent",
fmt.Sprintf("Processing message from %s:%s: %s", msg.Channel, msg.SenderID, logContent),
2026-02-20 18:03:11 +00:00
map[string]any{
"channel": msg.Channel,
"chat_id": msg.ChatID,
"sender_id": msg.SenderID,
"session_key": msg.SessionKey,
},
)
2026-03-01 08:31:04 +00:00
msg = al.transcribeAudioInMessage(ctx, msg)
// Route system messages to processSystemMessage
if msg.Channel == "system" {
return al.processSystemMessage(ctx, msg)
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
route, agent, routeErr := al.resolveMessageRoute(msg)
// Commands are checked before requiring a successful route.
// Global commands (/help, /show, /switch) work even when routing fails;
// context-dependent commands check their own Runtime fields and report
// "unavailable" when the required capability is nil.
if response, handled := al.handleCommand(ctx, msg, agent); handled {
return response, nil
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
if routeErr != nil {
return "", routeErr
}
// Reset message-tool state for this round so we don't skip publishing due to a previous round.
if tool, ok := agent.Tools.Get("message"); ok {
if resetter, ok := tool.(interface{ ResetSentInRound() }); ok {
resetter.ResetSentInRound()
}
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
// Resolve session key from route, while preserving explicit agent-scoped keys.
scopeKey := resolveScopeKey(route, msg.SessionKey)
sessionKey := scopeKey
logger.InfoCF("agent", "Routed message",
2026-02-20 18:03:11 +00:00
map[string]any{
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
"agent_id": agent.ID,
"scope_key": scopeKey,
"session_key": sessionKey,
"matched_by": route.MatchedBy,
"route_agent": route.AgentID,
"route_channel": route.Channel,
})
return al.runAgentLoop(ctx, agent, processOptions{
SessionKey: sessionKey,
Channel: msg.Channel,
ChatID: msg.ChatID,
UserMessage: msg.Content,
Media: msg.Media,
DefaultResponse: defaultResponse,
EnableSummary: true,
SendResponse: false,
})
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
func (al *AgentLoop) resolveMessageRoute(msg bus.InboundMessage) (routing.ResolvedRoute, *AgentInstance, error) {
route := al.registry.ResolveRoute(routing.RouteInput{
Channel: msg.Channel,
AccountID: inboundMetadata(msg, metadataKeyAccountID),
Peer: extractPeer(msg),
ParentPeer: extractParentPeer(msg),
GuildID: inboundMetadata(msg, metadataKeyGuildID),
TeamID: inboundMetadata(msg, metadataKeyTeamID),
})
agent, ok := al.registry.GetAgent(route.AgentID)
if !ok {
agent = al.registry.GetDefaultAgent()
}
if agent == nil {
return routing.ResolvedRoute{}, nil, fmt.Errorf("no agent available for route (agent_id=%s)", route.AgentID)
}
return route, agent, nil
}
func resolveScopeKey(route routing.ResolvedRoute, msgSessionKey string) string {
if msgSessionKey != "" && strings.HasPrefix(msgSessionKey, sessionKeyAgentPrefix) {
return msgSessionKey
}
return route.SessionKey
}
func (al *AgentLoop) processSystemMessage(
ctx context.Context,
msg bus.InboundMessage,
) (string, error) {
if msg.Channel != "system" {
return "", fmt.Errorf(
"processSystemMessage called with non-system message channel: %s",
msg.Channel,
)
}
logger.InfoCF("agent", "Processing system message",
2026-02-20 18:03:11 +00:00
map[string]any{
"sender_id": msg.SenderID,
"chat_id": msg.ChatID,
})
// Parse origin channel from chat_id (format: "channel:chat_id")
var originChannel, originChatID string
if idx := strings.Index(msg.ChatID, ":"); idx > 0 {
originChannel = msg.ChatID[:idx]
originChatID = msg.ChatID[idx+1:]
} else {
originChannel = "cli"
originChatID = msg.ChatID
}
// Extract subagent result from message content
// Format: "Task 'label' completed.\n\nResult:\n<actual content>"
content := msg.Content
if idx := strings.Index(content, "Result:\n"); idx >= 0 {
content = content[idx+8:] // Extract just the result part
}
// Skip internal channels - only log, don't send to user
if constants.IsInternalChannel(originChannel) {
logger.InfoCF("agent", "Subagent completed (internal channel)",
2026-02-20 18:03:11 +00:00
map[string]any{
2026-02-13 09:51:47 +00:00
"sender_id": msg.SenderID,
"content_len": len(content),
"channel": originChannel,
})
return "", nil
}
// Use default agent for system messages
agent := al.registry.GetDefaultAgent()
if agent == nil {
return "", fmt.Errorf("no default agent for system message")
}
// Use the origin session for context
sessionKey := routing.BuildAgentMainSessionKey(agent.ID)
return al.runAgentLoop(ctx, agent, processOptions{
SessionKey: sessionKey,
Channel: originChannel,
ChatID: originChatID,
UserMessage: fmt.Sprintf("[System: %s] %s", msg.SenderID, msg.Content),
DefaultResponse: "Background task completed.",
EnableSummary: false,
SendResponse: true,
})
}
// runAgentLoop is the core message processing logic.
func (al *AgentLoop) runAgentLoop(
ctx context.Context,
agent *AgentInstance,
opts processOptions,
) (string, error) {
// 0. Record last channel for heartbeat notifications (skip internal channels and cli)
if opts.Channel != "" && opts.ChatID != "" {
if !constants.IsInternalChannel(opts.Channel) {
channelKey := fmt.Sprintf("%s:%s", opts.Channel, opts.ChatID)
if err := al.RecordLastChannel(channelKey); err != nil {
logger.WarnCF(
"agent",
"Failed to record last channel",
map[string]any{"error": err.Error()},
)
}
}
}
// 1. Build messages (skip history for heartbeat)
var history []providers.Message
var summary string
if !opts.NoHistory {
history = agent.Sessions.GetHistory(opts.SessionKey)
summary = agent.Sessions.GetSummary(opts.SessionKey)
}
messages := agent.ContextBuilder.BuildMessages(
history,
summary,
opts.UserMessage,
opts.Media,
opts.Channel,
opts.ChatID,
2026-02-04 11:06:13 +00:00
)
// Resolve media:// refs to base64 data URLs (streaming)
maxMediaSize := al.cfg.Agents.Defaults.GetMaxMediaSize()
messages = resolveMediaRefs(messages, al.mediaStore, maxMediaSize)
// 2. Save user message to session
agent.Sessions.AddMessage(opts.SessionKey, "user", opts.UserMessage)
// 3. Run LLM iteration loop
finalContent, iteration, err := al.runLLMIteration(ctx, agent, messages, opts)
if err != nil {
return "", err
}
// If last tool had ForUser content and we already sent it, we might not need to send final response
// This is controlled by the tool's Silent flag and ForUser content
// 4. Handle empty response
if finalContent == "" {
finalContent = opts.DefaultResponse
}
// 5. Save final assistant message to session
agent.Sessions.AddMessage(opts.SessionKey, "assistant", finalContent)
agent.Sessions.Save(opts.SessionKey)
// 6. Optional: summarization
if opts.EnableSummary {
al.maybeSummarize(agent, opts.SessionKey, opts.Channel, opts.ChatID)
}
// 7. Optional: send response via bus
if opts.SendResponse {
al.bus.PublishOutbound(ctx, bus.OutboundMessage{
Channel: opts.Channel,
ChatID: opts.ChatID,
Content: finalContent,
})
}
// 8. Log response
responsePreview := utils.Truncate(finalContent, 120)
logger.InfoCF("agent", fmt.Sprintf("Response: %s", responsePreview),
2026-02-20 18:03:11 +00:00
map[string]any{
"agent_id": agent.ID,
"session_key": opts.SessionKey,
"iterations": iteration,
"final_length": len(finalContent),
})
return finalContent, nil
}
func (al *AgentLoop) targetReasoningChannelID(channelName string) (chatID string) {
if al.channelManager == nil {
return ""
}
if ch, ok := al.channelManager.GetChannel(channelName); ok {
return ch.ReasoningChannelID()
}
return ""
}
func (al *AgentLoop) handleReasoning(
ctx context.Context,
reasoningContent, channelName, channelID string,
) {
if reasoningContent == "" || channelName == "" || channelID == "" {
return
}
// Check context cancellation before attempting to publish,
// since PublishOutbound's select may race between send and ctx.Done().
if ctx.Err() != nil {
return
}
// Use a short timeout so the goroutine does not block indefinitely when
// the outbound bus is full. Reasoning output is best-effort; dropping it
// is acceptable to avoid goroutine accumulation.
pubCtx, pubCancel := context.WithTimeout(ctx, 5*time.Second)
defer pubCancel()
if err := al.bus.PublishOutbound(pubCtx, bus.OutboundMessage{
Channel: channelName,
ChatID: channelID,
Content: reasoningContent,
}); err != nil {
// Treat context.DeadlineExceeded / context.Canceled as expected
// (bus full under load, or parent canceled). Check the error
// itself rather than ctx.Err(), because pubCtx may time out
// (5 s) while the parent ctx is still active.
// Also treat ErrBusClosed as expected — it occurs during normal
// shutdown when the bus is closed before all goroutines finish.
if errors.Is(err, context.DeadlineExceeded) || errors.Is(err, context.Canceled) ||
errors.Is(err, bus.ErrBusClosed) {
logger.DebugCF("agent", "Reasoning publish skipped (timeout/cancel)", map[string]any{
"channel": channelName,
"error": err.Error(),
})
} else {
logger.WarnCF("agent", "Failed to publish reasoning (best-effort)", map[string]any{
"channel": channelName,
"error": err.Error(),
})
}
}
}
// runLLMIteration executes the LLM call loop with tool handling.
2026-02-20 18:03:11 +00:00
func (al *AgentLoop) runLLMIteration(
ctx context.Context,
agent *AgentInstance,
messages []providers.Message,
opts processOptions,
) (string, int, error) {
2026-02-04 11:06:13 +00:00
iteration := 0
var finalContent string
// Determine effective model tier for this conversation turn.
// selectCandidates evaluates routing once and the decision is sticky for
// all tool-follow-up iterations within the same turn so that a multi-step
// tool chain doesn't switch models mid-way through.
activeCandidates, activeModel := al.selectCandidates(agent, opts.UserMessage, messages)
for iteration < agent.MaxIterations {
2026-02-04 11:06:13 +00:00
iteration++
logger.DebugCF("agent", "LLM iteration",
2026-02-20 18:03:11 +00:00
map[string]any{
"agent_id": agent.ID,
"iteration": iteration,
"max": agent.MaxIterations,
})
// Build tool definitions
providerToolDefs := agent.Tools.ToProviderDefs()
2026-02-04 11:06:13 +00:00
// Log LLM request details
logger.DebugCF("agent", "LLM request",
2026-02-20 18:03:11 +00:00
map[string]any{
"agent_id": agent.ID,
"iteration": iteration,
"model": activeModel,
"messages_count": len(messages),
"tools_count": len(providerToolDefs),
"max_tokens": agent.MaxTokens,
"temperature": agent.Temperature,
"system_prompt_len": len(messages[0].Content),
})
// Log full messages (detailed)
logger.DebugCF("agent", "Full LLM request",
2026-02-20 18:03:11 +00:00
map[string]any{
"iteration": iteration,
"messages_json": formatMessagesForLog(messages),
"tools_json": formatToolsForLog(providerToolDefs),
})
// Call LLM with fallback chain if multiple candidates are configured.
var response *providers.LLMResponse
var err error
llmOpts := map[string]any{
"max_tokens": agent.MaxTokens,
"temperature": agent.Temperature,
"prompt_cache_key": agent.ID,
}
// parseThinkingLevel guarantees ThinkingOff for empty/unknown values,
// so checking != ThinkingOff is sufficient.
if agent.ThinkingLevel != ThinkingOff {
if tc, ok := agent.Provider.(providers.ThinkingCapable); ok && tc.SupportsThinking() {
llmOpts["thinking_level"] = string(agent.ThinkingLevel)
} else {
logger.WarnCF("agent", "thinking_level is set but current provider does not support it, ignoring",
map[string]any{"agent_id": agent.ID, "thinking_level": string(agent.ThinkingLevel)})
}
}
callLLM := func() (*providers.LLMResponse, error) {
if len(activeCandidates) > 1 && al.fallback != nil {
fbResult, fbErr := al.fallback.Execute(
ctx,
activeCandidates,
func(ctx context.Context, provider, model string) (*providers.LLMResponse, error) {
return agent.Provider.Chat(ctx, messages, providerToolDefs, model, llmOpts)
},
)
if fbErr != nil {
return nil, fbErr
}
if fbResult.Provider != "" && len(fbResult.Attempts) > 0 {
logger.InfoCF(
"agent",
fmt.Sprintf("Fallback: succeeded with %s/%s after %d attempts",
fbResult.Provider, fbResult.Model, len(fbResult.Attempts)+1),
map[string]any{"agent_id": agent.ID, "iteration": iteration},
)
}
return fbResult.Response, nil
}
return agent.Provider.Chat(ctx, messages, providerToolDefs, activeModel, llmOpts)
}
// Retry loop for context/token errors
maxRetries := 2
for retry := 0; retry <= maxRetries; retry++ {
response, err = callLLM()
if err == nil {
break
}
errMsg := strings.ToLower(err.Error())
// Check if this is a network/HTTP timeout — not a context window error.
isTimeoutError := errors.Is(err, context.DeadlineExceeded) ||
strings.Contains(errMsg, "deadline exceeded") ||
strings.Contains(errMsg, "client.timeout") ||
strings.Contains(errMsg, "timed out") ||
strings.Contains(errMsg, "timeout exceeded")
// Detect real context window / token limit errors, excluding network timeouts.
isContextError := !isTimeoutError && (strings.Contains(errMsg, "context_length_exceeded") ||
strings.Contains(errMsg, "context window") ||
strings.Contains(errMsg, "maximum context length") ||
strings.Contains(errMsg, "token limit") ||
strings.Contains(errMsg, "too many tokens") ||
strings.Contains(errMsg, "max_tokens") ||
strings.Contains(errMsg, "invalidparameter") ||
strings.Contains(errMsg, "prompt is too long") ||
strings.Contains(errMsg, "request too large"))
if isTimeoutError && retry < maxRetries {
backoff := time.Duration(retry+1) * 5 * time.Second
logger.WarnCF("agent", "Timeout error, retrying after backoff", map[string]any{
"error": err.Error(),
"retry": retry,
"backoff": backoff.String(),
})
time.Sleep(backoff)
continue
}
if isContextError && retry < maxRetries {
logger.WarnCF(
"agent",
"Context window error detected, attempting compression",
map[string]any{
"error": err.Error(),
"retry": retry,
},
)
if retry == 0 && !constants.IsInternalChannel(opts.Channel) {
al.bus.PublishOutbound(ctx, bus.OutboundMessage{
Channel: opts.Channel,
ChatID: opts.ChatID,
Content: "Context window exceeded. Compressing history and retrying...",
})
}
al.forceCompression(agent, opts.SessionKey)
newHistory := agent.Sessions.GetHistory(opts.SessionKey)
newSummary := agent.Sessions.GetSummary(opts.SessionKey)
messages = agent.ContextBuilder.BuildMessages(
newHistory, newSummary, "",
nil, opts.Channel, opts.ChatID,
)
continue
}
break
}
2026-02-04 11:06:13 +00:00
if err != nil {
logger.ErrorCF("agent", "LLM call failed",
2026-02-20 18:03:11 +00:00
map[string]any{
"agent_id": agent.ID,
"iteration": iteration,
"error": err.Error(),
})
return "", iteration, fmt.Errorf("LLM call failed after retries: %w", err)
2026-02-04 11:06:13 +00:00
}
go al.handleReasoning(
ctx,
response.Reasoning,
opts.Channel,
al.targetReasoningChannelID(opts.Channel),
)
logger.DebugCF("agent", "LLM response",
map[string]any{
"agent_id": agent.ID,
"iteration": iteration,
"content_chars": len(response.Content),
"tool_calls": len(response.ToolCalls),
"reasoning": response.Reasoning,
"target_channel": al.targetReasoningChannelID(opts.Channel),
"channel": opts.Channel,
})
// Check if no tool calls - we're done
2026-02-04 11:06:13 +00:00
if len(response.ToolCalls) == 0 {
finalContent = response.Content
logger.InfoCF("agent", "LLM response without tool calls (direct answer)",
2026-02-20 18:03:11 +00:00
map[string]any{
"agent_id": agent.ID,
"iteration": iteration,
"content_chars": len(finalContent),
})
2026-02-04 11:06:13 +00:00
break
}
normalizedToolCalls := make([]providers.ToolCall, 0, len(response.ToolCalls))
for _, tc := range response.ToolCalls {
normalizedToolCalls = append(normalizedToolCalls, providers.NormalizeToolCall(tc))
}
// Log tool calls
toolNames := make([]string, 0, len(normalizedToolCalls))
for _, tc := range normalizedToolCalls {
toolNames = append(toolNames, tc.Name)
}
logger.InfoCF("agent", "LLM requested tool calls",
2026-02-20 18:03:11 +00:00
map[string]any{
"agent_id": agent.ID,
"tools": toolNames,
"count": len(normalizedToolCalls),
"iteration": iteration,
})
// Build assistant message with tool calls
2026-02-04 11:06:13 +00:00
assistantMsg := providers.Message{
Role: "assistant",
Content: response.Content,
ReasoningContent: response.ReasoningContent,
2026-02-04 11:06:13 +00:00
}
for _, tc := range normalizedToolCalls {
2026-02-04 11:06:13 +00:00
argumentsJSON, _ := json.Marshal(tc.Arguments)
// Copy ExtraContent to ensure thought_signature is persisted for Gemini 3
extraContent := tc.ExtraContent
thoughtSignature := ""
if tc.Function != nil {
thoughtSignature = tc.Function.ThoughtSignature
}
2026-02-04 11:06:13 +00:00
assistantMsg.ToolCalls = append(assistantMsg.ToolCalls, providers.ToolCall{
ID: tc.ID,
Type: "function",
Name: tc.Name,
2026-02-04 11:06:13 +00:00
Function: &providers.FunctionCall{
Name: tc.Name,
Arguments: string(argumentsJSON),
ThoughtSignature: thoughtSignature,
2026-02-04 11:06:13 +00:00
},
ExtraContent: extraContent,
ThoughtSignature: thoughtSignature,
2026-02-04 11:06:13 +00:00
})
}
messages = append(messages, assistantMsg)
// Save assistant message with tool calls to session
agent.Sessions.AddFullMessage(opts.SessionKey, assistantMsg)
// Execute tool calls in parallel
type indexedAgentResult struct {
result *tools.ToolResult
tc providers.ToolCall
}
agentResults := make([]indexedAgentResult, len(normalizedToolCalls))
var wg sync.WaitGroup
for i, tc := range normalizedToolCalls {
agentResults[i].tc = tc
wg.Add(1)
go func(idx int, tc providers.ToolCall) {
defer wg.Done()
argsJSON, _ := json.Marshal(tc.Arguments)
argsPreview := utils.Truncate(string(argsJSON), 200)
logger.InfoCF("agent", fmt.Sprintf("Tool call: %s(%s)", tc.Name, argsPreview),
map[string]any{
"agent_id": agent.ID,
"tool": tc.Name,
"iteration": iteration,
})
// Create async callback for tools that implement AsyncExecutor.
// When the background work completes, this publishes the result
// as an inbound system message so processSystemMessage routes it
// back to the user via the normal agent loop.
asyncCallback := func(_ context.Context, result *tools.ToolResult) {
// Send ForUser content directly to the user (immediate feedback),
// mirroring the synchronous tool execution path.
if !result.Silent && result.ForUser != "" {
outCtx, outCancel := context.WithTimeout(context.Background(), 5*time.Second)
defer outCancel()
_ = al.bus.PublishOutbound(outCtx, bus.OutboundMessage{
Channel: opts.Channel,
ChatID: opts.ChatID,
Content: result.ForUser,
})
}
// Determine content for the agent loop (ForLLM or error).
content := result.ForLLM
if content == "" && result.Err != nil {
content = result.Err.Error()
}
if content == "" {
return
}
logger.InfoCF("agent", "Async tool completed, publishing result",
map[string]any{
"tool": tc.Name,
"content_len": len(content),
"channel": opts.Channel,
})
pubCtx, pubCancel := context.WithTimeout(context.Background(), 5*time.Second)
defer pubCancel()
_ = al.bus.PublishInbound(pubCtx, bus.InboundMessage{
Channel: "system",
SenderID: fmt.Sprintf("async:%s", tc.Name),
ChatID: fmt.Sprintf("%s:%s", opts.Channel, opts.ChatID),
Content: content,
})
}
toolResult := agent.Tools.ExecuteWithContext(
ctx,
tc.Name,
tc.Arguments,
opts.Channel,
opts.ChatID,
asyncCallback,
)
agentResults[idx].result = toolResult
}(i, tc)
}
wg.Wait()
// Process results in original order (send to user, save to session)
for _, r := range agentResults {
// Send ForUser content to user immediately if not Silent
if !r.result.Silent && r.result.ForUser != "" && opts.SendResponse {
al.bus.PublishOutbound(ctx, bus.OutboundMessage{
Channel: opts.Channel,
ChatID: opts.ChatID,
Content: r.result.ForUser,
})
logger.DebugCF("agent", "Sent tool result to user",
2026-02-20 18:03:11 +00:00
map[string]any{
"tool": r.tc.Name,
"content_len": len(r.result.ForUser),
})
}
// If tool returned media refs, publish them as outbound media
feat(feishu,tools): add outbound media delivery via send_file tool (#1156) * feat(feishu): implement SendMedia and add send_file tool Add outbound media support for the Feishu channel so the agent can send images and files to users via the MediaStore pipeline. Feishu channel: - SendMedia dispatches media parts as image or file uploads - sendImage uploads via Image.Create then sends image message - sendFile uploads via File.Create then sends file message - feishuFileType maps extensions to Feishu file_type values send_file tool: - New tool lets the LLM send a local file to the current chat - Validates path, registers file in MediaStore, returns media ref - Agent loop wires tool registration, MediaStore propagation, and context updates Tested on Radxa Cubie A7A (arm64) with Feishu websocket channel. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): publish outbound media regardless of SendResponse flag The SendResponse flag controls whether the agent loop publishes the final text response (callers that publish it themselves set this to false). However, the media publish path was also gated behind this flag, which meant tool-produced media was silently dropped for normal channel messages. Media should be published immediately when a tool returns media refs, independent of how the text response is delivered. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(tools): use magic-bytes MIME detection and add file size limit to send_file - Replace hardcoded extension-to-MIME map with h2non/filetype (magic bytes) + mime.TypeByExtension fallback, consistent with the vision pipeline in resolveMediaRefs - Add configurable max file size check (defaults to config.DefaultMaxMediaSize, 20 MB) to prevent oversized uploads - Add tests for magic-bytes detection, extension fallback, size limit, and default max size Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): add ForEachTool to AgentRegistry for cross-agent tool lookup Extract the pattern of iterating agents to find a named tool into AgentRegistry.ForEachTool, simplifying SetMediaStore propagation. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent,tools): adapt send_file to ctx-based channel injection after upstream refactor Replace ContextualTool interface (removed upstream) with direct ctx reading in SendFileTool.Execute, using ToolChannel/ToolChatID helpers. Remove updateToolContexts which is no longer needed since ExecuteWithContext already injects channel/chatID into ctx for all tools. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(tools): support toggling send_file tool via config Add SendFileConfig with Enabled field to ToolsConfig, defaulting to true. Wrap send_file tool registration in loop.go with the config check, consistent with the pattern used by other tools. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 11:42:52 +00:00
if len(r.result.Media) > 0 {
parts := make([]bus.MediaPart, 0, len(r.result.Media))
for _, ref := range r.result.Media {
part := bus.MediaPart{Ref: ref}
if al.mediaStore != nil {
if _, meta, err := al.mediaStore.ResolveWithMeta(ref); err == nil {
part.Filename = meta.Filename
part.ContentType = meta.ContentType
part.Type = inferMediaType(meta.Filename, meta.ContentType)
}
}
parts = append(parts, part)
}
al.bus.PublishOutboundMedia(ctx, bus.OutboundMediaMessage{
Channel: opts.Channel,
ChatID: opts.ChatID,
Parts: parts,
})
}
// Determine content for LLM based on tool result
contentForLLM := r.result.ForLLM
if contentForLLM == "" && r.result.Err != nil {
contentForLLM = r.result.Err.Error()
2026-02-04 11:06:13 +00:00
}
toolResultMsg := providers.Message{
Role: "tool",
Content: contentForLLM,
ToolCallID: r.tc.ID,
2026-02-04 11:06:13 +00:00
}
messages = append(messages, toolResultMsg)
// Save tool result message to session
agent.Sessions.AddFullMessage(opts.SessionKey, toolResultMsg)
}
2026-02-04 11:06:13 +00:00
}
return finalContent, iteration, nil
}
// selectCandidates returns the model candidates and resolved model name to use
// for a conversation turn. When model routing is configured and the incoming
// message scores below the complexity threshold, it returns the light model
// candidates instead of the primary ones.
//
// The returned (candidates, model) pair is used for all LLM calls within one
// turn — tool follow-up iterations use the same tier as the initial call so
// that a multi-step tool chain doesn't switch models mid-way.
func (al *AgentLoop) selectCandidates(
agent *AgentInstance,
userMsg string,
history []providers.Message,
) (candidates []providers.FallbackCandidate, model string) {
if agent.Router == nil || len(agent.LightCandidates) == 0 {
return agent.Candidates, agent.Model
}
_, usedLight, score := agent.Router.SelectModel(userMsg, history, agent.Model)
if !usedLight {
logger.DebugCF("agent", "Model routing: primary model selected",
map[string]any{
"agent_id": agent.ID,
"score": score,
"threshold": agent.Router.Threshold(),
})
return agent.Candidates, agent.Model
}
logger.InfoCF("agent", "Model routing: light model selected",
map[string]any{
"agent_id": agent.ID,
"light_model": agent.Router.LightModel(),
"score": score,
"threshold": agent.Router.Threshold(),
})
return agent.LightCandidates, agent.Router.LightModel()
}
// maybeSummarize triggers summarization if the session history exceeds thresholds.
func (al *AgentLoop) maybeSummarize(agent *AgentInstance, sessionKey, channel, chatID string) {
newHistory := agent.Sessions.GetHistory(sessionKey)
tokenEstimate := al.estimateTokens(newHistory)
threshold := agent.ContextWindow * agent.SummarizeTokenPercent / 100
if len(newHistory) > agent.SummarizeMessageThreshold || tokenEstimate > threshold {
summarizeKey := agent.ID + ":" + sessionKey
if _, loading := al.summarizing.LoadOrStore(summarizeKey, true); !loading {
go func() {
defer al.summarizing.Delete(summarizeKey)
logger.Debug("Memory threshold reached. Optimizing conversation history...")
al.summarizeSession(agent, sessionKey)
}()
}
}
}
// forceCompression aggressively reduces context when the limit is hit.
// It drops the oldest 50% of messages (keeping system prompt and last user message).
func (al *AgentLoop) forceCompression(agent *AgentInstance, sessionKey string) {
history := agent.Sessions.GetHistory(sessionKey)
if len(history) <= 4 {
return
}
// Keep system prompt (usually [0]) and the very last message (user's trigger)
// We want to drop the oldest half of the *conversation*
// Assuming [0] is system, [1:] is conversation
conversation := history[1 : len(history)-1]
if len(conversation) == 0 {
return
}
// Helper to find the mid-point of the conversation
mid := len(conversation) / 2
// New history structure:
// 1. System Prompt (with compression note appended)
// 2. Second half of conversation
// 3. Last message
droppedCount := mid
keptConversation := conversation[mid:]
newHistory := make([]providers.Message, 0, 1+len(keptConversation)+1)
// Append compression note to the original system prompt instead of adding a new system message
// This avoids having two consecutive system messages which some APIs (like Zhipu) reject
2026-02-20 18:03:11 +00:00
compressionNote := fmt.Sprintf(
"\n\n[System Note: Emergency compression dropped %d oldest messages due to context limit]",
droppedCount,
)
enhancedSystemPrompt := history[0]
enhancedSystemPrompt.Content = enhancedSystemPrompt.Content + compressionNote
newHistory = append(newHistory, enhancedSystemPrompt)
newHistory = append(newHistory, keptConversation...)
newHistory = append(newHistory, history[len(history)-1]) // Last message
// Update session
agent.Sessions.SetHistory(sessionKey, newHistory)
agent.Sessions.Save(sessionKey)
2026-02-20 18:03:11 +00:00
logger.WarnCF("agent", "Forced compression executed", map[string]any{
"session_key": sessionKey,
"dropped_msgs": droppedCount,
"new_count": len(newHistory),
})
}
// GetStartupInfo returns information about loaded tools and skills for logging.
2026-02-20 18:03:11 +00:00
func (al *AgentLoop) GetStartupInfo() map[string]any {
info := make(map[string]any)
agent := al.registry.GetDefaultAgent()
if agent == nil {
return info
}
// Tools info
toolsList := agent.Tools.List()
2026-02-20 18:03:11 +00:00
info["tools"] = map[string]any{
"count": len(toolsList),
"names": toolsList,
}
// Skills info
info["skills"] = agent.ContextBuilder.GetSkillsInfo()
// Agents info
2026-02-20 18:03:11 +00:00
info["agents"] = map[string]any{
"count": len(al.registry.ListAgentIDs()),
"ids": al.registry.ListAgentIDs(),
}
return info
}
// formatMessagesForLog formats messages for logging
func formatMessagesForLog(messages []providers.Message) string {
if len(messages) == 0 {
return "[]"
}
var sb strings.Builder
sb.WriteString("[\n")
for i, msg := range messages {
fmt.Fprintf(&sb, " [%d] Role: %s\n", i, msg.Role)
if len(msg.ToolCalls) > 0 {
sb.WriteString(" ToolCalls:\n")
for _, tc := range msg.ToolCalls {
fmt.Fprintf(&sb, " - ID: %s, Type: %s, Name: %s\n", tc.ID, tc.Type, tc.Name)
if tc.Function != nil {
fmt.Fprintf(
&sb,
" Arguments: %s\n",
utils.Truncate(tc.Function.Arguments, 200),
)
}
}
}
if msg.Content != "" {
content := utils.Truncate(msg.Content, 200)
fmt.Fprintf(&sb, " Content: %s\n", content)
}
if msg.ToolCallID != "" {
fmt.Fprintf(&sb, " ToolCallID: %s\n", msg.ToolCallID)
}
sb.WriteString("\n")
}
sb.WriteString("]")
return sb.String()
}
// formatToolsForLog formats tool definitions for logging
func formatToolsForLog(toolDefs []providers.ToolDefinition) string {
if len(toolDefs) == 0 {
return "[]"
}
var sb strings.Builder
sb.WriteString("[\n")
for i, tool := range toolDefs {
fmt.Fprintf(&sb, " [%d] Type: %s, Name: %s\n", i, tool.Type, tool.Function.Name)
fmt.Fprintf(&sb, " Description: %s\n", tool.Function.Description)
if len(tool.Function.Parameters) > 0 {
fmt.Fprintf(
&sb,
" Parameters: %s\n",
utils.Truncate(fmt.Sprintf("%v", tool.Function.Parameters), 200),
)
}
}
sb.WriteString("]")
return sb.String()
}
// summarizeSession summarizes the conversation history for a session.
func (al *AgentLoop) summarizeSession(agent *AgentInstance, sessionKey string) {
ctx, cancel := context.WithTimeout(context.Background(), 120*time.Second)
defer cancel()
history := agent.Sessions.GetHistory(sessionKey)
summary := agent.Sessions.GetSummary(sessionKey)
// Keep last 4 messages for continuity
if len(history) <= 4 {
return
}
toSummarize := history[:len(history)-4]
// Oversized Message Guard
maxMessageTokens := agent.ContextWindow / 2
validMessages := make([]providers.Message, 0)
omitted := false
for _, m := range toSummarize {
if m.Role != "user" && m.Role != "assistant" {
continue
}
msgTokens := len(m.Content) / 2
if msgTokens > maxMessageTokens {
omitted = true
continue
}
validMessages = append(validMessages, m)
}
if len(validMessages) == 0 {
return
}
// Multi-Part Summarization
var finalSummary string
if len(validMessages) > 10 {
mid := len(validMessages) / 2
part1 := validMessages[:mid]
part2 := validMessages[mid:]
s1, _ := al.summarizeBatch(ctx, agent, part1, "")
s2, _ := al.summarizeBatch(ctx, agent, part2, "")
2026-02-20 18:03:11 +00:00
mergePrompt := fmt.Sprintf(
"Merge these two conversation summaries into one cohesive summary:\n\n1: %s\n\n2: %s",
s1,
s2,
)
resp, err := agent.Provider.Chat(
ctx,
[]providers.Message{{Role: "user", Content: mergePrompt}},
nil,
agent.Model,
map[string]any{
fix: cache system prompt with mtime-based auto-invalidation (#607) Avoid rebuilding the entire system prompt on every BuildMessages() call by caching the static portion (identity, bootstrap, skills summary, memory) and only recomputing it when workspace source files change. Key changes: - ContextBuilder caches the static prompt behind an RWMutex with double-checked locking. Source file changes are detected via cheap os.Stat mtime checks so no explicit invalidation is needed. - Track file existence at cache time (existedAtCache map) so that newly created or deleted bootstrap/memory files also trigger a rebuild — the old modifiedSince() silently returned false on os.IsNotExist. - Walk the skills directory recursively with filepath.WalkDir to catch content-only edits at any nesting depth; directory mtime alone misses in-place file modifications on most filesystems. - ToolRegistry.sortedToolNames() sorts tool names before iteration, ensuring deterministic tool definition order across calls — a prerequisite for LLM-side prefix/KV cache reuse. - Merge all context (static + dynamic + summary) into a single system message for provider compatibility: the Anthropic adapter extracts messages[0] as the top-level system parameter, and Codex reads only the first system message as instructions. - Fix a data race in BuildMessages() where cachedSystemPrompt was read without holding the lock in a debug log statement. - Add tests: single system message invariant, mtime auto-invalidation, new-file creation detection, skill file content change, explicit InvalidateCache, cache stability, concurrent access (20 goroutines x 50 iterations, passes go test -race), and a benchmark.
2026-02-25 02:34:54 +00:00
"max_tokens": 1024,
"temperature": 0.3,
"prompt_cache_key": agent.ID,
2026-02-20 18:03:11 +00:00
},
)
if err == nil {
finalSummary = resp.Content
} else {
finalSummary = s1 + " " + s2
}
} else {
finalSummary, _ = al.summarizeBatch(ctx, agent, validMessages, summary)
}
if omitted && finalSummary != "" {
finalSummary += "\n[Note: Some oversized messages were omitted from this summary for efficiency.]"
}
if finalSummary != "" {
agent.Sessions.SetSummary(sessionKey, finalSummary)
agent.Sessions.TruncateHistory(sessionKey, 4)
agent.Sessions.Save(sessionKey)
}
}
// summarizeBatch summarizes a batch of messages.
2026-02-20 18:03:11 +00:00
func (al *AgentLoop) summarizeBatch(
ctx context.Context,
agent *AgentInstance,
batch []providers.Message,
existingSummary string,
) (string, error) {
var sb strings.Builder
sb.WriteString(
"Provide a concise summary of this conversation segment, preserving core context and key points.\n",
)
if existingSummary != "" {
sb.WriteString("Existing context: ")
sb.WriteString(existingSummary)
sb.WriteString("\n")
}
sb.WriteString("\nCONVERSATION:\n")
for _, m := range batch {
fmt.Fprintf(&sb, "%s: %s\n", m.Role, m.Content)
}
prompt := sb.String()
2026-02-20 18:03:11 +00:00
response, err := agent.Provider.Chat(
ctx,
[]providers.Message{{Role: "user", Content: prompt}},
nil,
agent.Model,
map[string]any{
fix: cache system prompt with mtime-based auto-invalidation (#607) Avoid rebuilding the entire system prompt on every BuildMessages() call by caching the static portion (identity, bootstrap, skills summary, memory) and only recomputing it when workspace source files change. Key changes: - ContextBuilder caches the static prompt behind an RWMutex with double-checked locking. Source file changes are detected via cheap os.Stat mtime checks so no explicit invalidation is needed. - Track file existence at cache time (existedAtCache map) so that newly created or deleted bootstrap/memory files also trigger a rebuild — the old modifiedSince() silently returned false on os.IsNotExist. - Walk the skills directory recursively with filepath.WalkDir to catch content-only edits at any nesting depth; directory mtime alone misses in-place file modifications on most filesystems. - ToolRegistry.sortedToolNames() sorts tool names before iteration, ensuring deterministic tool definition order across calls — a prerequisite for LLM-side prefix/KV cache reuse. - Merge all context (static + dynamic + summary) into a single system message for provider compatibility: the Anthropic adapter extracts messages[0] as the top-level system parameter, and Codex reads only the first system message as instructions. - Fix a data race in BuildMessages() where cachedSystemPrompt was read without holding the lock in a debug log statement. - Add tests: single system message invariant, mtime auto-invalidation, new-file creation detection, skill file content change, explicit InvalidateCache, cache stability, concurrent access (20 goroutines x 50 iterations, passes go test -race), and a benchmark.
2026-02-25 02:34:54 +00:00
"max_tokens": 1024,
"temperature": 0.3,
"prompt_cache_key": agent.ID,
2026-02-20 18:03:11 +00:00
},
)
if err != nil {
return "", err
}
return response.Content, nil
}
// estimateTokens estimates the number of tokens in a message list.
// Uses a safe heuristic of 2.5 characters per token to account for CJK and other
// overheads better than the previous 3 chars/token.
func (al *AgentLoop) estimateTokens(messages []providers.Message) int {
totalChars := 0
for _, m := range messages {
totalChars += utf8.RuneCountInString(m.Content)
}
// 2.5 chars per token = totalChars * 2 / 5
return totalChars * 2 / 5
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
func (al *AgentLoop) handleCommand(
ctx context.Context,
msg bus.InboundMessage,
agent *AgentInstance,
) (string, bool) {
if !commands.HasCommandPrefix(msg.Content) {
return "", false
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
if al.cmdRegistry == nil {
return "", false
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
rt := al.buildCommandsRuntime(agent)
executor := commands.NewExecutor(al.cmdRegistry, rt)
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
var commandReply string
result := executor.Execute(ctx, commands.Request{
Channel: msg.Channel,
ChatID: msg.ChatID,
SenderID: msg.SenderID,
Text: msg.Content,
Reply: func(text string) error {
commandReply = text
return nil
},
})
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
switch result.Outcome {
case commands.OutcomeHandled:
if result.Err != nil {
return mapCommandError(result), true
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
if commandReply != "" {
return commandReply, true
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
return "", true
default: // OutcomePassthrough — let the message fall through to LLM
return "", false
}
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
func (al *AgentLoop) buildCommandsRuntime(agent *AgentInstance) *commands.Runtime {
rt := &commands.Runtime{
Config: al.cfg,
ListAgentIDs: al.registry.ListAgentIDs,
ListDefinitions: al.cmdRegistry.Definitions,
GetEnabledChannels: func() []string {
if al.channelManager == nil {
return nil
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
return al.channelManager.GetEnabledChannels()
},
SwitchChannel: func(value string) error {
if al.channelManager == nil {
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
return fmt.Errorf("channel manager not initialized")
}
if _, exists := al.channelManager.GetChannel(value); !exists && value != "cli" {
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
return fmt.Errorf("channel '%s' not found or not enabled", value)
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
return nil
},
}
if agent != nil {
rt.GetModelInfo = func() (string, string) {
return agent.Model, al.cfg.Agents.Defaults.Provider
}
rt.SwitchModel = func(value string) (string, error) {
oldModel := agent.Model
agent.Model = value
return oldModel, nil
}
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
return rt
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
func mapCommandError(result commands.ExecuteResult) string {
if result.Command == "" {
return fmt.Sprintf("Failed to execute command: %v", result.Err)
}
return fmt.Sprintf("Failed to execute /%s: %v", result.Command, result.Err)
}
// extractPeer extracts the routing peer from the inbound message's structured Peer field.
func extractPeer(msg bus.InboundMessage) *routing.RoutePeer {
if msg.Peer.Kind == "" {
return nil
}
peerID := msg.Peer.ID
if peerID == "" {
if msg.Peer.Kind == "direct" {
peerID = msg.SenderID
} else {
peerID = msg.ChatID
}
}
return &routing.RoutePeer{Kind: msg.Peer.Kind, ID: peerID}
}
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
func inboundMetadata(msg bus.InboundMessage, key string) string {
if msg.Metadata == nil {
return ""
}
return msg.Metadata[key]
}
// extractParentPeer extracts the parent peer (reply-to) from inbound message metadata.
func extractParentPeer(msg bus.InboundMessage) *routing.RoutePeer {
feat(commands): centralized command registry with sub-command routing (#959) * feat(commands): Session management [Phase 1/2] command centralization and registration * docs: add design for command registry post-review fixes Documents the architecture decisions for fixing 5 Important issues from code review: SubCommand pattern, Deps struct, command-group files, Executor caching, and Telegram registration dedup. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add SubCommand type and EffectiveUsage method Introduce SubCommand struct for declaring sub-commands structurally within a parent command Definition. The EffectiveUsage() method auto-generates usage strings from sub-command names and args, preventing drift between help text and actual handler behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add Deps struct and secondToken helper, remove dead contains() Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat(commands): add sub-command routing to Executor Uses Registry.Lookup for O(1) command dispatch instead of iterating all definitions. Definitions with SubCommands are routed to matching sub-command handlers. Missing or unknown sub-commands reply with auto-generated usage. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): split into command-group files with Deps injection Extract show/list/start/help into individual cmd_*.go files. Replace config.Config parameter with Deps struct for runtime data. Restore /show agents and /list agents sub-commands. Use EffectiveUsage for auto-generated help text. Bridge external callers (agent/loop.go, telegram.go) with Deps wrapper until Task 5 fully wires the Deps fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * perf(commands): cache Executor in AgentLoop, wire Deps with runtime callbacks Create Executor once in NewAgentLoop instead of per-message. Deps closures capture AgentLoop pointer for late-bound access to channelManager and runtime agent model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(telegram): remove duplicate initBotCommands, keep async startCommandRegistration only Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore(commands): restore Outcome comments and annotate Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): consolidate /switch into commands package, fix ! prefix Move /switch model and /switch channel handling from inline loop.go logic into cmd_switch.go using the SubCommand + Deps pattern. This removes the OutcomePassthrough branch in handleCommand entirely. Also replace the hardcoded "/" prefix check with commands.HasCommandPrefix so that "!" prefixed commands are correctly routed to the Executor. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * chore: add docs/plans to .gitignore and untrack existing files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings - Remove dead ExecuteResult.Reply field and unused branch in loop.go - Extract shared agentsHandler for /show agents and /list agents - Remove redundant firstToken/secondToken (use nthToken instead) - Simplify Telegram startup: pass BuiltinDefinitions directly - Centralize req.Reply nil guard in executeDefinition - Extract unavailableMsg constant (was duplicated 5 times) - Remove unused MessageID from Request - Remove stale "reserved for Phase 2" comment on Deps.Config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): replace Deps with per-request Runtime Separate stateless Registry (cached on AgentLoop) from per-request Runtime (passed to handlers at execution time). This enables future session management features to inject per-request context without modifying the command registry. - Rename Deps → Runtime, move to runtime.go - Change Handler signature: func(ctx, req) error → func(ctx, req, rt *Runtime) error - NewExecutor now takes (registry, runtime) — executor is created per-request - BuiltinDefinitions() no longer takes parameters (stateless) - AgentLoop caches cmdRegistry, builds Runtime via buildRuntime() - Update all cmd_*.go handlers and tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * style: fix gci import grouping and godoc formatting Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(onboard): skip legacy AGENT.md when copying embedded workspace templates The workspace/ directory contains both AGENT.md (legacy) and AGENTS.md (current). copyEmbeddedToTarget was copying both, causing the test TestCopyEmbeddedToTargetUsesAgentsMarkdown to fail. Skip AGENT.md during the walk to match the expected behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(agent): address self-review comments on loop.go - Move cmdRegistry init into struct literal (review comment #11) - Rename buildRuntime → buildCommandsRuntime for clarity (review comment #12) - Add comment to default switch case explaining passthrough (review comment #13) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): address code review findings on naming and correctness - Rename dispatcher.go → request.go (no Dispatcher type remains) - Rename cmd_agents.go → handler_agents.go (shared handler, not a top-level command) - Add modelMu to protect AgentInstance.Model writes in SwitchModel - Add ListDefinitions to Runtime so /help uses registry instead of BuiltinDefinitions() - Fix SwitchChannel message: validation-only callback should not say "Switched" - Propagate Reply errors in executor instead of discarding with _ = - Add HasCommandPrefix unit test Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(onboard): extract legacy filename to constant Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(agent): handle commands before route error check Move handleCommand() before the routeErr gate so global commands (/help, /show, /switch) remain available even when routing fails. Context-dependent commands that need a routed agent will report "unavailable" through their nil-Runtime guards. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * revert: remove unnecessary AGENT.md skip in onboard Reverts 02d0c04 and 74deae1. The test failure was caused by a local leftover workspace/AGENT.md file (gitignored but embedded by go:embed). Deleting the local file fixes the root cause; the code-level skip was never needed. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: executeDefinition Unknown option * fix(agent): use routed agent for model commands, restore Telegram command diff - Remove modelMu: message processing is serial, no concurrent writes - Pass routed agent to handleCommand/buildCommandsRuntime instead of always using default agent - GetModelInfo/SwitchModel are nil when agent is nil (route failed), handlers reply "unavailable" - Restore GetMyCommands + slices.Equal check before SetMyCommands to avoid unnecessary Telegram API calls on restart Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(commands): remove unintended config mutation in SwitchModel SwitchModel should only update the routed agent's runtime Model field. Writing to cfg.Agents.Defaults.ModelName was a behavioral change that corrupts the default agent config when switching a non-default agent. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor(commands): move /switch channel to /check channel /switch channel only validates availability, not actually switching. Rename to /check channel to match actual behavior. /switch channel now shows a redirect message pointing users to the new command. Addresses review feedback from yinwm on PR #959. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-06 09:31:40 +00:00
parentKind := inboundMetadata(msg, metadataKeyParentPeerKind)
parentID := inboundMetadata(msg, metadataKeyParentPeerID)
if parentKind == "" || parentID == "" {
return nil
}
return &routing.RoutePeer{Kind: parentKind, ID: parentID}
}