Context

Diagnoses and fixes failing tests, lint errors, or build issues. Reads the error output and applies targeted fixes.

Inputs

$ARGUMENTS - Optional: "tests", "lint", "build", or specific error message. Defaults to detecting the issue.

Steps

1. Identify the Problem

If $ARGUMENTS specifies the type:

•tests → Run the test suite and capture failures
•lint → Run the linter and capture errors
•build → Run the build and capture errors

If no argument, detect the most likely issue:

•Run build first (build errors block everything)
•Then run tests
•Then run linter

2. Analyze Errors

For each error:

•Read the error message and stack trace
•Identify the file and line number
•Read the surrounding code to understand context
•Check learnings/*.md for known patterns that might apply

3. Apply Fixes

Fix errors one at a time, starting with the most fundamental:

Build errors:

•Missing imports/usings
•Type mismatches
•Missing method implementations
•Dependency issues

Test failures:

•Incorrect assertions
•Missing mock setup
•Changed API contracts
•Stale test data

Lint errors:

•Formatting issues
•Naming convention violations
•Unused imports
•Style violations

4. Verify Fixes

After each fix:

•Re-run the specific failing command
•Confirm the error is resolved
•Check that no new errors were introduced

5. Report

code

## Fix Report

**Issues found:** X
**Issues fixed:** Y

### Fixes Applied
1. **[file:line]** [What was wrong] → [What was fixed]
2. ...

### Remaining Issues (if any)
- [Issue that could not be auto-fixed and why]

Verification

•All originally failing tests/lint/build now pass
•No new errors introduced by the fixes
•Fixes follow project standards