The go-nostr RelayConnect returns immediately after the WebSocket
handshake, but the relay has not yet sent the AUTH challenge. Calling
relay.Auth() right away signs the auth event with an empty challenge
tag, which the Buzz relay rejects permanently ('verification failed'
then 'authentication already failed' on retries).
Add a 2s sleep between RelayConnect and Auth() to give the background
read loop time to receive and store the challenge. The relay typically
delivers it within a few hundred milliseconds.
Implements a Buzz channel backed by a Nostr relay:
- kind:9 events scoped by "h" tag, "p" tags for mentions
- NIP-42 (kind:22242) auth before subscription
- optional threaded replies via "e" tag
- registered as config.ChannelBuzz with BuzzSettings
Two bugs prevented the usage block from ever reaching the wire:
1. CallLLM read turnStateFromContext(ctx), but the raw ctx is not seeded
with the turn state (only turnCtx is), so SetLastUsage/SetLastFinishReason
were dropped — GetLastUsage() returned nil at finalize. Set them on the
ts parameter directly, which is also what the streaming publisher reads.
2. The manager wraps the channel streamer in finalizeHookStreamer /
splitMarkerStreamer, neither of which forwarded SetTurnUsage (it is not
part of the bus.Streamer interface), so the type assertion in the
publisher's Finalize failed silently. Mirror the existing SetModelName
forwarding: add a turnUsageStreamer interface + setStreamerTurnUsage
helper and SetTurnUsage methods on both wrappers (splitMarker also stores
and re-applies usage to each freshly-begun part streamer).
Adds regression tests asserting both wrappers forward SetTurnUsage.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
When handling inbound Telegram messages from forum topics, InboundContext.ChatID was set to the plain chat ID (without thread). This caused outbound messages to lose the thread routing, sending replies to the General topic instead of the specific forum topic — even though typing indicators correctly targeted the right thread because StartTyping received the compositeChatID.
The fix sets InboundContext.ChatID to compositeChatID (chatID/threadID format) for forum messages, matching the format already used for session routing and typing indicators. TopicID is still preserved as a redundant fallback for any path that strips the composite format.
Closes#3110
Add defer-recover to 11 goroutines across 4 files to prevent
ungoroutine panics from crashing the entire process:
- pkg/tools/toolloop.go: parallel tool execution
- pkg/channels/manager.go: HTTP server (x2), channel registration
- pkg/events/subscription.go: concurrent dispatch, timeout handler,
watchContext
- pkg/tools/shell.go: cmd.Wait, PTY cmd.Wait, PTY read, pipe read
Key design decisions:
- Recover handlers send fallback values to channels (shell done,
subscription done) to prevent deadlocks when the producer panics
- PTY cmd.Wait sets session.Status='error' on panic for consistency
- toolloop sets ErrorResult on panic so the LLM gets a meaningful
response instead of a nil result
- subscription.go uses log.Printf to match existing invokeHandler style
- Other files use project logger (ErrorCF) with stack traces
Refs: FIX-PLAN-0.3.0 #2
Replace silently discarded json.Marshal and json.Unmarshal errors with
explicit checks. If serialization fails, log a warning and either
return early (for the config-level marshal/unmarshal) or skip the
channel (for per-channel marshal). This prevents silent data loss
when channel configuration contains unexpected types.
errutil.go: Change %v to %w in ClassifySendError and ClassifyNetError so callers can use errors.Is/errors.As on the underlying HTTP/network error.
isolated_command_transport.go: Change %v to %w in Close() and Write() error paths for the same reason.
Add 3 tests covering scenarios that previously panicked: 1) missing enabled key in settings 2) enabled field with non-bool type 3) teams_webhook with webhooks using map[string]any from JSON unmarshal
Address remaining review feedback: 1) Add HistoryTokens field to ContextUsage/ContextStats, showing history-only token count in /context and frontend UI alongside SummarizeAtTokens so users can see the actual summarization trigger comparison. 2) Remove .codebuddy/github-contribute/ state files accidentally included in the PR.
Two type assertions in toChannelHashes could panic when channel config values had unexpected types from JSON unmarshal: 1) value[enabled].(bool) panics if the key is missing or not a bool 2) vv.(map[string]string) panics when JSON unmarshal produces map[string]any. Add ok checks to safely handle both cases.
When an incoming group message is received, the inbound context ChatID was set to the raw group number without the group: prefix. This caused the outbound reply to use send_private_msg instead of send_group_msg. Fix by using the prefixed chatID as inbound context ChatID. Closes#3002
The SDK renamed ReceiveIdTypeChatId to CreateMessageV1ReceiveIDTypeChatId
in v3.9.4. Update all 5 usages in feishu_64.go and bump the dependency
version.
This fixes the build failure for Dependabot PR #3005.
The /context command previously showed only the hard budget compression
threshold (contextWindow - maxTokens), which confused users who expected
to see the soft summarization trigger from summarize_token_percent.
This commit adds SummarizeAtTokens alongside the existing CompressAtTokens
so that both thresholds are visible:
- Compress at: contextWindow - maxTokens (hard budget, triggers proactive
compression when exceeded)
- Summarize at: contextWindow * summarizeTokenPercent / 100 (soft trigger,
matches maybeSummarize's threshold)
The fix updates the /context command output, the Web UI popover, and the
pico channel WebSocket payload.
Fixes#2968
The auxiliary message filtering introduced in #2892 incorrectly drops
tool_calls messages when there is an active stream or a tombstone from
a recently finished stream.
This causes tool_calls to not be delivered to the UI when users make
consecutive requests, as the second request's tool_calls message arrives
while the first request's stream is still active or within the 30-second
tombstone window.
The fix excludes tool_calls from auxiliary message filtering, since they
represent new tool invocations that must be delivered to the UI, not
stale auxiliary content like feedback or thoughts.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Discord only downloaded audio attachments before passing them to the agent. Non-audio attachments (images, videos, files) were passed as raw Discord CDN URLs, which do not flow through resolveMediaRefs and are not serialized as vision inputs.
Download every attachment, store it in the MediaStore with Discord's filename and content type metadata, and emit a media placeholder tag that matches the attachment kind. This lets resolveMediaRefs replace the placeholder with the local path-bearing tag and encode supported images for vision-capable providers. If a download fails, keep the previous raw URL fallback.
* feat: add request-scoped context policies
Add named turn profiles under agents.defaults so callers can opt into
per-request context and tool policies without changing default chat behavior.
Profiles can disable history, system context, skill prompts, or tools, and can
limit skills/tools with allow lists. Wire profile selection through Pico message
payloads, agent turn execution, Web chat selection, and Web visual config.
Reject invalid turn profiles before saving config through Web APIs and document
the new request context policy behavior.
* fix: address turn profile review blockers
* feat: simplify request context policy config
* fix: suppress tool prompt when turn tools are disabled
* fix: enforce turn profile tool restrictions
* feat(chat,seahorse): persist and display model_name across history
* test(seahorse): fix lint regressions in repair coverage
* fix(pico): preserve model_name in live updates
* fix(pico): preserve model_name through live stream wrappers
