Agent skill

fixme-todo-cleanup

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/fixme-todo-cleanup

SKILL.md

FIXME/TODO Cleanup Skill

Field Value
Date 2026-01-01
Objective Systematically resolve all FIXME/TODO items in a directory
Outcome 7 PRs merged, 3 issues closed
Category debugging

When to Use

  • User requests cleanup of FIXME/TODO items across a codebase
  • Technical debt reduction initiative
  • Pre-release code quality sweep
  • Issue tracking consolidation

Verified Workflow

1. Discovery Phase

bash
# Find all FIXME/TODO items
grep -rn "FIXME\|TODO" shared/ --include="*.mojo"

2. Categorization

Group items by:

  • Stale references: FIXME pointing to closed issues
  • Missing implementations: Placeholder code needing real logic
  • External blockers: Items blocked by language/compiler features
  • Documentation: TODO in docs describing future work

3. Execution Pattern

For each actionable item:

bash
# 1. Create branch from latest main
git checkout main && git pull origin main
git checkout -b <issue-number>-<description>

# 2. Make the fix (one FIXME per commit)

# 3. Validate
just pre-commit-all
pixi run mojo test tests/

# 4. Commit with issue reference
git add . && git commit -m "fix(scope): description

Closes #<issue>"

# 5. Push and create PR with auto-merge
git push -u origin <branch-name>
gh pr create --body "Closes #<issue>"
gh pr merge --auto --rebase

4. Handle Flaky CI

bash
# Retry failed jobs
gh run rerun <run-id> --failed

Failed Attempts

1. @value to @fieldwise_init Migration

What was tried: Replace deprecated @value decorator with @fieldwise_init

Error:

mojo
error: 'TrainingCallbacks' has an explicitly declared fieldwise initializer

Why it failed: @fieldwise_init generates an __init__ method, conflicting with custom __init__

Solution: Remove decorator entirely, keep explicit Copyable, Movable conformances:

mojo
# Before (broken)
@fieldwise_init
struct TrainingCallbacks(Copyable, Movable):
    fn __init__(out self, verbose: Bool = True):
        self.verbose = verbose

# After (working)
struct TrainingCallbacks(Copyable, Movable):
    var verbose: Bool

    fn __init__(out self, verbose: Bool = True):
        self.verbose = verbose

2. Ralph-Loop Hook Stuck

What was tried: Set active: false in ralph-loop.local.md

Why it failed: Hook continued firing despite deactivation flag

Solution: Delete the file entirely:

bash
rm .claude/ralph-loop.local.md

Results & Parameters

Items Resolved

PR Issue Description
#3035 - Update stale issue references
#3036 #3031 Remove MXFP4 FIXME
#3037 #3033 Implement conftest fixtures
#3038 #3034 Implement script_runner
#3039 #3034 Implement dataset_loaders
#3040 #3034 Export script utilities
#3041 #3032 Update ExTensor Array API

Items NOT Resolved (External Blockers)

File Line Blocker
profiling.mojo 650 Mojo FileIO stability
logging.mojo 442 Mojo env var support
mixed_precision.mojo 284, 368 Compiler SIMD FP16 support
training/__init__.mojo 412 Track 4 Python data loader
trainer_interface.mojo 392 Track 4 Python data loader

Git Configuration

yaml
branch_naming: "<issue-number>-<description>"
commit_format: "type(scope): description\n\nCloses #<issue>"
merge_strategy: "rebase"
auto_merge: true

Key Learnings

  1. Categorize first: Not all FIXME/TODO items are actionable - some are blocked by external dependencies
  2. One PR per fix: Keeps changes atomic and reviewable
  3. Auto-merge is essential: Reduces manual intervention for straightforward fixes
  4. Retry flaky tests: Some CI tests are non-deterministic - retry before investigating
  5. Mojo decorator conflicts: @fieldwise_init cannot coexist with custom \_\_init\_\_

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results