/fix-issue

Fix a bug from error output, logs, or user description.

Usage

/fix-issue <error message or bug description>

Workflow

1. •

   Find or create issue (MANDATORY — do this BEFORE reading code or editing anything): Search open beads issues for a match. If none found, create one with bd create --title="[Bug] ..." --type=bug --priority=2. Claim it with bd update <id> --status=in_progress. Do NOT proceed to step 2 without an in_progress issue.

2. •

   Search closed issues for context: Use a sub-agent to search closed beads issues for prior fixes to similar problems:

   code

   Task(subagent_type="Explore", prompt="Search closed beads issues for context about <error/bug>.
   Run: bd list --status=closed | grep -i <keyword>
   Then bd show each relevant match.
   Return concise summary: relevant files, patterns, key decisions.")
   

3. •

   Locate the problem: Use error messages, stack traces, and closed issue context to find the relevant source files. Read them to understand the bug.

4. •

   Implement the fix: Make the minimal change needed. Follow project conventions (error wrapping, slog logging, constructor injection).

5. •

   Run tests: Execute make test to verify the fix doesn't break anything. If the fix area lacks test coverage, add a test case.

6. •

   Close the issue: bd close <id> --reason="Fixed: <what was wrong and what was changed>"