Skip to content

fix(ui): do not count no-data guidance as failed chat steps - #353

Merged
vishr merged 1 commit into
mainfrom
fix/chat-guidance-not-failures
Oct 18, 2026
Merged

vishr merged 1 commit into
mainfrom
fix/chat-guidance-not-failures

Conversation

@vishr

@vishr vishr commented Oct 18, 2026 •

Copy link
Copy Markdown
Member

Summary

The agent's own rejected drafts (for example a query_telemetry panel it then corrects) still count as failed steps. Hiding those reliably needs a distinct server error code for validation rejections; a name- or target-matching heuristic hid real failures in review and was dropped.

Verification

  • New regression (fails without the fix): guidance results do not count; a real failure in the same turn still shows.
  • 59 chat tests, UI lint, just ui, just ui-check.
  • Codex adversarial review of the guidance exclusion.
  • just check in the pre-push hook.

Since empty results became no_data and unknown_service guidance, chat
counted them as failed steps, so a successful investigation showed "2 steps
failed" while the agent was only told which signal exists. Guidance results
are no longer listed as failures; other tool errors still are.

Claude-Session: https://claude.ai/code/session_018HfyEXHyGD1utmKF4CQtyQ
@vishr
vishr force-pushed the fix/chat-guidance-not-failures branch from 7f9bf6f to d4677a8 Compare October 18, 2026 23:06
@vishr
vishr merged commit dd0be95 into main Oct 18, 2026
16 checks passed
@vishr
vishr deleted the fix/chat-guidance-not-failures branch October 18, 2026 23:23
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant