How this command is triggered — by the user, by Claude, or both
Slash command
/workflows:fix <description>The summary Claude sees in its command listing — used to decide when to auto-load this command
# Fix Workflow Investigate and resolve bugs, errors, and issues. ## Usage ## Workflow Chain ### Phase 1: Investigation (sleuth) Spawn the **sleuth** agent to investigate: **sleuth** will: - Search for error patterns - Trace code execution - Check logs and stack traces - Identify root cause ### Phase 2: Implementation (kraken) Once cause is identified, spawn **kraken** to fix: ### Phase 3: Verification (arbiter) Spawn **arbiter** to verify: ## Options | Option | Effect | |--------|--------| | `--dry-run` | Investigate only, don't fix | | `--no-test` | Skip test verificati...
Investigate and resolve bugs, errors, and issues.
/fix "login fails silently"
/fix "API returns 500 on POST"
/fix hook "pre-commit not running"
Spawn the sleuth agent to investigate:
Use Task tool:
subagent_type: "sleuth"
prompt: "Investigate: $DESCRIPTION
1. Reproduce the issue if possible
2. Find the root cause
3. Identify affected code
4. Report findings with confidence level"
sleuth will:
Once cause is identified, spawn kraken to fix:
Use Task tool:
subagent_type: "kraken"
prompt: "Fix the issue identified by sleuth:
Root cause: [from sleuth]
Affected files: [from sleuth]
Implement the fix following TDD:
1. Write a failing test that reproduces the bug
2. Implement the minimal fix
3. Verify test passes"
Spawn arbiter to verify:
Use Task tool:
subagent_type: "arbiter"
prompt: "Verify the fix:
1. Run all related tests
2. Check for regressions
3. Verify the original issue is resolved
4. Report results"
| Option | Effect |
|---|---|
--dry-run | Investigate only, don't fix |
--no-test | Skip test verification |
--no-commit | Don't auto-commit the fix |
After completion, summarize:
npx claudepluginhub sethdford/claude-toolkit --plugin workflows/fix-issueFixes GitHub issue by number using parallel root cause analysis, hypothesis testing, similar issue detection, fixes, tests, and prevention recommendations.
/debugRuns structured debugging with parallel agents: traces bugs via phases, diagnoses root causes with evidence, proposes minimal fixes, executes after approval. Tracks in tasks/todo.md.
/debugDebugs and fixes bugs end-to-end: checks context and known issues, reproduces, investigates root cause, test-first fix, and creates PR.
/fixIteratively repairs code errors until zero remain via autonomous loop, applying one atomic fix per iteration with auto-revert on failure. Supports --target, --scope, --category, --iterations flags.
/fixFixes issues automatically by locating root cause, applying minimal changes, and running tests. Also supports systematic debug mode (/fix debug) for analysis without code changes.
/fixApplies auto-fixable code review findings: dispatches fix-group agents, post-reviews working tree, commits survivors, reverts regressions, updates artifact.