Skip to content

feat: Harden Agent tool observability with lifecycle records, safe payloads, and OTel spans#409

Open
eryue0220 wants to merge 3 commits into
heymrun:mainfrom
eryue0220:feat/agent-tool-o11y
Open

feat: Harden Agent tool observability with lifecycle records, safe payloads, and OTel spans#409
eryue0220 wants to merge 3 commits into
heymrun:mainfrom
eryue0220:feat/agent-tool-o11y

Conversation

@eryue0220

Copy link
Copy Markdown
Contributor

Summary

  1. Normalize Agent tool_calls with stable tool_call_id, lifecycle status (success / error / pending / timeout / cancelled), timestamps, and aggregate tool_metrics, without changing execution results or breaking historical traces.
  2. Sanitize persisted tool request/response payloads (redaction + size/depth limits); keep live execution and HITL resume state untouched.
  3. Emit opt-in OpenTelemetry child spans heym.agent.tool.execute under the Agent node span, with tool identity/status attributes and no raw args/results by default.
  4. Surface pending/timeout/cancelled and tool metrics in Traces/Chat UI, and update docs/tests accordingly.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Development

Successfully merging this pull request may close these issues.

2 participants