Claude AutoFix

Version updated for https://github.com/enriconunes/claude-autofix-action to version v1.4.

• This action is used across all versions by ? repositories.

Action Type

This is a Composite action.

Go to the GitHub Marketplace to find the latest changes.

Action Summary

The Claude AutoFix GitHub Action automates the process of identifying and resolving failing Python tests in pull requests. It diagnoses test failures, posts a detailed comment explaining the issue, and uses Claude AI to generate a code fix, which is then submitted as a new pull request for review. This action simplifies troubleshooting and accelerates the bug-fixing workflow by reducing manual intervention in identifying and correcting errors.

Release notes

Reduced the analysis comment posted on PRs from a verbose multi-section report to a compact 4-line summary per failing test.

Before: long output with Root Cause Analysis, Recommended Fix, unified diff, Post-Fix Expectations, and PR guidance — often redundant and repetitive.

After: File: functions/dividir.py | Line: 2 Error: expected 5, got 8 Cause: The function uses subtraction (-) instead of division (/). Fix: change return a - b to return a / b

This also reduces token usage per analysis call.

• enriconunes
• GitHub Actions