Skip to content

fix(tui): refresh working status when tools finish - #225

Merged
liuedcson merged 1 commit into
mainfrom
fix/tool-running-delay-20261009
Oct 9, 2026
Merged

liuedcson merged 1 commit into
mainfrom
fix/tool-running-delay-20261009

Conversation

@MelodyVAR

Copy link
Copy Markdown
Collaborator

When a tool finishes before the next assistant output, the Step working indicator keeps showing Running..., Reading..., or Editing.... The tool result and spinner are already complete, but the tool-verb hold remains set, so a slow next response looks like a tool that is still running.

Release that hold and clear the previous reasoning phase on tool_execution_end, then refresh the indicator after stopping the completed tool's spinner. Any remaining parallel tool keeps its own active verb; otherwise the indicator returns to Working... while preserving elapsed time and token counts.

This fixes the reproducible status display issue. The separately reported 30–60 second execution delay was not reproduced and is not claimed fixed by this PR.

Validation:

  • Seven new event-to-render regression cases cover successful and failed run/read/edit completion, a simulated minute without further model output, and switching to a remaining parallel tool.
  • All 41 tests in the seven related UI test files pass.
  • npm run check passes.
  • ./test.sh passes on Node 22.19.0, including the offline build and all workspace test suites (credential-dependent tests are skipped by the isolated runner).

@liuedcson
liuedcson merged commit 56e24b8 into main Oct 9, 2026
2 of 3 checks passed
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.

2 participants