* fix: remove redundant tools definitions from system prompt
Tools are already provided to the LLM via JSON schema through
ToProviderDefs(), so the text-based tools section in the system
prompt is redundant.
This removes the buildToolsSection() logic and the tools field
from ContextBuilder, reducing system prompt length while maintaining
the "ALWAYS use tools" rule reminder.
Fixes#731
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* fix: correct spelling 'initialized' (was 'initialised')
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Whitespace is a linter that checks for unnecessary newlines at the start and end of functions, if, for, etc.
Signed-off-by: Kai Xia <kaix+github@fastmail.com>
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.
The spawn tool accepts empty strings as valid task arguments, which
causes a subagent to run with no meaningful work. The subagent's
completion message is then routed back to the originating channel
(e.g. Signal, Discord), where the main agent processes it and may
hallucinate an unrelated response that gets sent to users.
Validate that the task parameter is non-empty after trimming whitespace.
Related: #545
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* cleanup dead code.
Signed-off-by: Kai Xia <kaix+github@fastmail.com>
* add these two back with flag.
Signed-off-by: Kai Xia <kaix+github@fastmail.com>
* fix ci
Signed-off-by: Kai Xia <kaix+github@fastmail.com>
* remove this confusing line
Signed-off-by: Kai Xia <kaix+github@fastmail.com>
* make fmt
Signed-off-by: Kai Xia <kaix+github@fastmail.com>
* remove unused method.
picked up by golangci-lint run
Signed-off-by: Kai Xia <kaix+github@fastmail.com>
---------
Signed-off-by: Kai Xia <kaix+github@fastmail.com>
Walk backwards over preceding tool messages to find the nearest assistant
with ToolCalls, instead of only checking the immediate predecessor. Add
unit tests for sanitizeHistoryForProvider covering key edge cases.
* chore: Update default host bindings from 0.0.0.0 to 127.0.0.1 for various services and examples.
* config: Update default host bindings to 0.0.0.0 for improved Docker accessibility and add related documentation.
* chore: resolve conflict
* chore: remove link
* docs: Add a tip for Docker users regarding gateway host configuration to the French and Vietnamese READMEs.
* fix: typo issue
* docs: Update Chinese README.zh.md.
- Update Dockerfile to use golang:1.25-alpine to match go.mod (go 1.25.7)
- Optimize logger by avoiding string concatenation in file writes
- Add explicit empty string assignment for fieldStr when no fields
These changes improve build consistency and reduce memory allocations
in the hot logging path, which is important for the project's goal
of running on resource-constrained devices (<10MB RAM).
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
* chore: Update default host bindings from 0.0.0.0 to 127.0.0.1 for various services and examples.
* config: Update default host bindings to 0.0.0.0 for improved Docker accessibility and add related documentation.
* refactor: reimplement filesystem tools with `os.OpenRoot` for enhanced security and simplified path validation.
* chore: revert other PR content from this branch
* docs: Update Chinese README.
* docs: Update Chinese README.
* docs: Update Chinese README.
* refactor: Reorder filesystem helper functions, extract directory entry formatting logic, and enhance `WriteFileTool`'s result message.
* feat: Enhance `mkdirAllInRoot` to prevent creating directories over existing files and add tests for directory creation functionality.
* Refactor filesystem tools to use a `fileReadWriter` interface for both host and sandboxed I/O, improving atomic writes and error handling.
* refactor: unify filesystem read/write operations with atomic write guarantees and clearer naming.
* refactor: rename `appendFileWithRW` function to `appendFile`
* refactor: unify filesystem access by introducing a `fileSystem` interface and updating tools to use it directly, removing `os.Root` dependency from `sandboxFs`.
* chore: run make fmt
* fix: `validatePath` now returns an error when the workspace is empty.
The configuration field for specifying the model has been renamed from
"model" to "model_name" for better clarity and consistency with the
model_list configuration.
A GetModelName() accessor method has been added to maintain backward
compatibility. Existing configurations using the old "model" field will
continue to work correctly.
This change affects:
- Configuration structure (AgentDefaults struct)
- All references across the codebase
- Documentation in all language variants
- Example configuration files
* feat: integrate Tavily search
* fix: set include_raw_content to false in Tavily search as wealready get relevant data inside content
* refactor: update Go type declarations to `any`, apply formatting fixes.
Add comprehensive unit tests for the ToolRegistry covering registration,
lookup, execution, context injection, async callbacks, schema generation,
provider definition conversion, and concurrent access.
Fix a defensive edge case in Truncate where a negative maxLen would cause
a slice bounds panic, and add table-driven tests covering boundary
conditions, zero/negative lengths, and Unicode handling.
Co-authored-by: Cursor <cursoragent@cursor.com>
Mistral's API strictly validates tool_calls in assistant messages and
rejects non-standard fields. The ToolCall struct had Name and Arguments
as top-level JSON fields, duplicating data already in Function.Name
and Function.Arguments. OpenAI silently ignored these extras but
Mistral returns 422.
Change json tags to "-" so these internal fields are no longer
serialized to API payloads while remaining available in Go code.
Add Mistral as a first-class provider alongside the 17 existing ones.
Mistral uses the OpenAI-compatible API at https://api.mistral.ai/v1
with provider-specific model prefix stripping (mistral/model → model).
Changes:
- Add Mistral to ProvidersConfig, IsEmpty(), HasProvidersConfig()
- Add mistral entry in default model_list (defaults.go)
- Add mistral protocol in factory_provider.go and getDefaultAPIBase()
- Add mistral prefix stripping in openai_compat normalizeModel()
- Add mistral case in legacy factory.go resolveProviderSelection()
- Add mistral migration entry in ConvertProvidersToModelList()
- Add mistral to supported providers in migrate/config.go
- Add mistral section in config.example.json
- Update AllProviders test (17 → 18 providers)
Tested end-to-end with mistral-small-latest model.
Models like Moonshot kimi-k2.5 and DeepSeek-R1 return a
reasoning_content field in assistant messages. When thinking is enabled,
the API requires this field to be echoed back in subsequent requests.
PicoClaw was silently dropping it, causing 400 errors on tool-call
round-trips.
- Add ReasoningContent to Message and LLMResponse types
- Parse reasoning_content in openai_compat parseResponse()
- Carry reasoning_content through assistant tool-call messages
- Add unit test for reasoning_content parsing
Fixes#588
* feat(discord): add mention_only option for @-mention responses
Add MentionOnly config option to Discord channel. When enabled, the bot
only responds when explicitly @-mentioned, useful for shared servers.
- Add MentionOnly bool field to DiscordConfig
- Store botUserID on startup for mention checking
- Check m.Mentions before processing messages when MentionOnly is true
- Update config example and README documentation
* fix(discord): resolve race condition and strip mention from content
- Get botUserID before opening session to avoid race condition
- Add stripBotMention to remove @mention from message content
- Handles both <@USER_ID> and <@!USER_ID> mention formats
* fix(discord): skip mention_only check for DMs
DMs should always be responded to regardless of mention_only setting.
Added check to skip the mention_only logic when GuildID is empty.
* Update README.md
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
---------
Co-authored-by: Hua Audio <161028864+Huaaudio@users.noreply.github.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* fix(onebot): add metadata for direct and group message handling
* fix(qq): add metadata for direct and group message handling
* fix(dingtalk): add metadata for direct and group message handling
* fix(feishu): add metadata for direct and group message handling
* fix(whatsapp): add metadata for direct and group message handlinga
* fix(line): add metadata for direct and group message handling
* fix(maixcam): add metadata for person detection handling
* fix(config): add default session configuration with DMScope
Resolved conflicts:
- pkg/config/config.go: Removed duplicate DefaultConfig() (already in defaults.go)
- pkg/config/defaults.go: Updated Temperature to *float64 (nil default)
Upstream changes included:
- Temperature changed from float64 to *float64 (nil means use provider default)
- New HeartbeatConfig and DevicesConfig
- Various agent and tool improvements
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace all claude-sonnet-4 references with claude-sonnet-4.6 across
codebase including documentation, tests, and configuration examples.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Remove duplicate model_name check in ValidateModelList to support
load balancing feature where multiple configs can share the same
model_name for round-robin selection.
Update tests to reflect the new behavior.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add github_copilot to the supportedProviders map to match
the providers handled in MergeConfig.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Remove sync.RWMutex and rrCounters from Config struct
- Simplify GetModelConfig to use global atomic counter for load balancing
- Remove unnecessary locks from HasProvidersConfig, SaveConfig, etc.
- Add buildModelWithProtocol helper to handle models with existing prefix
- Fix TestCreateProviderReturnsHTTPProviderForOpenRouter to use model_list
- Upgrade all Claude 3 references to Claude 4 across documentation
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Auth fixes:
- Fix OpenAI/Anthropic OAuth and token login to update ModelList
- Fix logout to clear AuthMethod in ModelList
- Add helper functions: isOpenAIModel, isAnthropicModel, isAntigravityModel
- Fix slice bounds panic in isAntigravityModel using strings.HasPrefix
- All auth operations now preserve existing model_list configuration
Factory provider fixes:
- Add OAuth support for openai protocol in CreateProviderFromConfig
- CodexAuthProvider is now used when auth_method is oauth/token
Default model updates:
- OpenAI login: set default model to gpt-5.2
- Anthropic login: set default model to claude-sonnet-4
- Antigravity login: set default model to gemini-flash (remove provider field)
Model changes:
- Change default OpenAI model from gpt-4o to gpt-5.2
- gpt-5.2 is compatible with Codex API (chatgpt.com backend)
- Update all README files, config examples, and migration code
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Include all 17 supported providers in default config as templates
- Each entry has model_name, model, api_base, and empty api_key
- Add comments with API key links for each provider
- Keep onboard message simple (only OpenRouter and Ollama)
- Fix duplicate model_name (cerebras-llama-3.3-70b)
Providers included:
Zhipu, OpenAI, Anthropic, DeepSeek, Gemini, Qwen, Moonshot,
Groq, OpenRouter, NVIDIA, Cerebras, Volcengine, ShengsuanYun,
Antigravity, GitHub Copilot, Ollama, VLLM
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: add MaxTokens and Temperature fields to AgentInstance and update related logic
* feat: add MaxTokens and Temperature options to SubagentManager and update tool loop logic
* feat: add default temperature handling and update related tests
* feat: allow temperature 0 and distinguish unset
* fix: format MockLLMProvider struct in subagent_tool_test.go
Update model configurations to use provider-specific protocols (zhipu, vllm,
gemini, shengsuanyun, deepseek, volcengine) instead of using the generic
"openai" protocol for all providers. This change ensures each provider
uses its correct protocol identifier and model naming convention.
Add support for persisting thought_signature metadata from Google/Gemini 3
models. This introduces ExtraContent and GoogleExtra types to handle
provider-specific metadata, and ensures thought signatures are properly
preserved through the tool call lifecycle.
- Move provider creation logic to factory_provider.go with protocol-based approach
- Add OpenAIProviderConfig with WebSearch support and embedded ProviderConfig
- Add maxTokensField to OpenAI-compatible provider for configurable token field
- Introduce new providers: Ollama, DeepSeek, GitHubCopilot, Antigravity, Qwen
- Remove redundant CreateProvider function from factory.go
- Add ThoughtSignature field to FunctionCall for tool response handling
- Remove duplicate Name field assignment in tool loop
- Update tests to reflect new provider configuration structure
Append emergency compression note to the original system prompt
instead of creating a separate system message. Some APIs like
Zhipu reject two consecutive system messages.
* fix: keep Discord typing indicator alive during agent processing
Discord's ChannelTyping() expires after ~10s, but agent processing
(LLM + tool execution) typically takes 30-60s+. Replace single-fire
ChannelTyping() with a self-managed typing loop inside DiscordChannel.
- startTyping(chatID): goroutine refreshes ChannelTyping every 8s
- stopTyping(chatID): called in Send() when response is dispatched
- Stop() cleans up all typing goroutines on shutdown
- startTyping placed after all early returns to prevent goroutine leaks
Typing lifecycle fully contained in channel layer, no interface changes.
Fixes#390
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: add goroutine safety to Discord typing indicator
- Add 5-minute timeout as safety net to prevent indefinite goroutine leaks
when agent produces no outbound message (empty response, panic, etc.)
- Listen on c.ctx.Done() so goroutine exits when channel context is cancelled
- Log ChannelTyping() errors at debug level for diagnostics (rate limits, session closed)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: replace bool map with set-style map for internal channels
Use map[string]struct{} and comma-ok idiom for clearer and more idiomatic membership checks.
* Update pkg/constants/channels.go
Co-authored-by: Harsh Bansal <122075346+harshbansal7@users.noreply.github.com>
---------
Co-authored-by: Harsh Bansal <122075346+harshbansal7@users.noreply.github.com>
* fix: change BotStatus type to json.RawMessage and add isAPIResponse function
* feat(onebot): add rich media, API callback, keepalive and voice transcription
Comprehensive improvements to the OneBot channel for better NapCatQQ
compatibility:
- Add echo-based API callback mechanism (sendAPIRequest) for
request/response correlation via pending map
- Add WebSocket ping/pong keepalive (30s ping, 60s read deadline)
- Fetch bot self ID via get_login_info on connect/reconnect
- Refactor parseMessageContentEx into parseMessageSegments supporting
image, record, video, file, reply, face, forward segments
- Add voice transcription via Groq transcriber (SetTranscriber)
- Switch to message segment array format for sending with auto reply
quote via lastMessageID tracking
- Add message_sent event handling and detailed notice event processing
(recall, poke, group increase/decrease, friend add, etc.)
- Use sync/atomic for echoCounter, optimize listen() lock pattern
- Clean up pending callbacks on Stop(), defer temp file cleanup
- Mount Groq transcriber on OneBot channel in main.go gateway
* feat(onebot): add user ID allowlist check for incoming messages
- Currently, the agent does not respond to messages sent by users outside the allowlist.
* refactor(onebot): simplify channel implementation and add emoji reaction
- onebot.go from 1179 to 980 lines (~17%)
When no provider field is set but model is specified, use the user's model
as ModelName for the first provider. This maintains backward compatibility
with old configs that relied on implicit provider selection and ensures
GetModelConfig(model) can find the model by its configured name.
Add validation to ensure model_name is unique across all entries in
model_list. This prevents potential conflicts when multiple model
configs share the same model_name identifier.
- Preserve user's configured model during config migration (issue #5)
- Simplify ExtractProtocol using strings.Cut
- Extract NormalizeToolCall to shared utility, removing ~70 lines of duplicate code
- Clean up unused fields in providerMigrationConfig struct
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
