Add code-simplifier skill with three-aspect review

[xingyaoww]

• A human has tested these changes.

---

Why

The code-simplifier plugin from the Claude Plugins Official repository provides useful code simplification capabilities, but it is a single monolithic pass and references Claude-specific conventions (CLAUDE.md). This PR migrates and adapts it for OpenHands as a proper skill with progressive disclosure, split into three focused review dimensions.

Summary

• Add new code-simplifier skill that analyzes recently modified code across three dimensions: code reuse, code quality, and efficiency
• Each dimension has its own detailed reference document (references/code-reuse-review.md, references/code-quality-review.md, references/efficiency-review.md) for progressive disclosure
• SKILL.md supports sub-agent delegation (parallel review) with fallback to sequential review
• All CLAUDE.md references replaced with AGENTS.md for OpenHands compatibility
• Registered in marketplaces/openhands-extensions.json and README catalog updated via sync script

How to Test

1. Verify the skill structure: ls -R skills/code-simplifier/
2. Verify sync passes: pip install pyyaml && python scripts/sync_extensions.py --check
3. Verify marketplace registration: grep -A 8 'code-simplifier' marketplaces/openhands-extensions.json
4. Review the SKILL.md for correct frontmatter, description, and references to AGENTS.md
5. Confirm each reference doc has tagged output format ([REUSE], [QUALITY], [EFFICIENCY])

Notes

• Adapted from https://github.com/anthropics/claude-plugins-official/blob/main/plugins/code-simplifier/agents/code-simplifier.md
• The skill uses keyword triggers (simplify, refine, cleanup) rather than slash commands
• Word counts: SKILL.md ~640 words, reference docs ~430–620 words each — well within progressive disclosure guidelines

This PR was created by an AI agent (OpenHands) on behalf of the user.

@xingyaoww can click here to continue refining the PR

[all-hands-bot]

🟡 Acceptable - Well-designed skill with good progressive disclosure structure, but uses deprecated trigger convention.

[IMPROVEMENT OPPORTUNITIES]

• Trigger format uses deprecated slash convention instead of keywords
• Scope specification could be more explicit for user clarity

[RISK ASSESSMENT]

⚠️ Risk Assessment: 🟢 LOW

New skill addition with no changes to existing code. Well-structured progressive disclosure, properly registered in marketplace, follows repository conventions (except trigger format). Limited blast radius - only affects users who invoke this specific skill.

VERDICT: ✅ Worth merging after addressing trigger convention

KEY INSIGHT: The three-aspect split (reuse, quality, efficiency) with separate reference documents is a pragmatic improvement over monolithic review - each dimension has clear, actionable criteria without forcing special-case handling.

[xingyaoww]

Addressed feedback in 5d8be69. Ready for another look.

[all-hands-bot]

🟢 Good taste - Well-designed skill addition with clear structure, pragmatic philosophy, and proper progressive disclosure.

[RISK ASSESSMENT]

⚠️ Risk Assessment: 🟢 LOW

New skill addition with no changes to existing code. Well-structured progressive disclosure with three focused review dimensions (reuse, quality, efficiency), properly registered in marketplace, follows repository conventions including the new punctuation style. Previous review feedback has been addressed.

VERDICT:

✅ Worth merging: Elegant solution that addresses a real need without over-engineering.

KEY INSIGHT:

The three-aspect breakdown with separate reference files demonstrates good taste in progressive disclosure - simple enough to understand quickly, detailed enough to be useful in practice.