reflection-loop

Self-correction via critique loop - 18.5 percentage point improvement

170 stars

byMiosa-osa

View on GitHub Installation ↓

Best use case

reflection-loop is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Self-correction via critique loop - 18.5 percentage point improvement

Teams using reflection-loop should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

You only need a quick one-off answer and do not need a reusable workflow.
You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/reflection/SKILL.md --create-dirs "https://raw.githubusercontent.com/Miosa-osa/canopy/main/library/skills/ai-patterns/reflection/SKILL.md"

Manual Installation

Download SKILL.md from GitHub
Place it in .claude/skills/reflection/SKILL.md inside your project
Restart your AI agent — it will auto-discover the skill

How reflection-loop Compares

Feature / Agent	reflection-loop	Standard Approach
Platform Support	Not specified	Limited / Varies
Context Awareness	High	Baseline
Installation Complexity	Unknown	N/A

Frequently Asked Questions

What does this skill do?

Self-correction via critique loop - 18.5 percentage point improvement

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

SKILL.md Source

# Reflection Loop Pattern

Self-correction mechanism that achieves 18.5 percentage point improvement (78.6% → 97.1%).

## Activation

This skill auto-activates on:
- Complex task completion
- Error detection
- Quality concerns
- User correction patterns

## Process

### 1. Initial Response
Generate first-pass response to the task.

### 2. Self-Critique
Ask yourself:
- Is the logic correct?
- Are there edge cases missed?
- Does this fully address the requirements?
- What could go wrong?
- Is the code/solution secure?

### 3. Identify Issues
List specific issues found:
- [ ] Logic errors
- [ ] Missing edge cases
- [ ] Security vulnerabilities
- [ ] Performance issues
- [ ] Incomplete requirements

### 4. Revise
For each issue:
1. Explain the problem
2. Propose fix
3. Apply fix
4. Verify improvement

### 5. Final Verification
- Re-check all criteria
- Confirm no regressions
- Validate requirements met

## Integration

The reflection loop integrates with:
- Code review workflow
- Test-driven development
- Verification before completion
- Learning engine (captures patterns)

## Metrics

Track reflection effectiveness:
- Issues caught in critique
- Revisions made
- Final quality score
- User corrections avoided

---

*Based on research showing 18.5pp improvement with reflection patterns*

## Command