Assistant incorrectly reports rejected tool action as completed

Resolved 💬 3 comments Opened Apr 4, 2026 by bsabri Closed Apr 8, 2026

Preflight Checklist

  • [x] I have searched existing issues and this hasn't been reported yet
  • [x] This is a single bug report (please file separate reports for different bugs)
  • [x] I am using the latest version of Claude Code

What's Wrong?

When a Bash tool call (e.g., pkill) is rejected by the user, the assistant's next message incorrectly describes the action as having been executed. Example: user rejects a pkill command, assistant responds with "The background tests were killed but memory hasn't
been freed yet" — implying the kill succeeded when it didn't.

Expected: Assistant should acknowledge the rejection and not describe the rejected action as completed.

What Should Happen?

Assistant should acknowledge the rejection and not describe the rejected action as completed

Error Messages/Logs

Steps to Reproduce

  1. Start a long-running background process (e.g., sleep 1000 &)
  2. Ask the assistant to kill it
  3. When the assistant issues a pkill or kill Bash command, reject/deny the tool call
  4. Observe the assistant's next message — it will describe the process as killed despite the rejection

Actual behavior: Assistant says something like "Killed. The process is no longer running..." even though the tool call was rejected.

Expected behavior: Assistant should say "The kill command was rejected. The process may still be running. Would you like me to try again?"

Claude Model

None

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.1.92 (Claude Code)

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

_No response_

View original on GitHub ↗

This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗