Best Claude Code Skills
obra/superpowers is the top-scored Claude Code skill framework (83/100, A-tier), on the strength of trigger design and process discipline. anthropics/skills is the canonical reference collection, and garrytan/gstack has the widest workflow coverage. Scored 2026-07-07, methodology v1.0.
The ranking
| # | Entry | Overall | Tier | Reliability | Leverage | Last commit |
|---|---|---|---|---|---|---|
| 1 | obra/superpowers | 83 | A-Tier | 84 | 90 | 2026-07-06 |
| 2 | anthropics/skills | 81 | A-Tier | 78 | 82 | 2026-07-01 |
| 3 | garrytan/gstack | 80 | A-Tier | 80 | 86 | 2026-06-25 |
| 4 | K-Dense-AI/scientific-agent-skills | 77 | A-Tier | 70 | 84 | 2026-07-06 |
| 5 | davila7/claude-code-templates | 75 | A-Tier | 70 | 78 | 2026-07-07 |
| 6 | wshobson/agents | 72 | A-Tier | 68 | 76 | 2026-07-06 |
| 7 | SuperClaude-Org/SuperClaude_Framework | 68 | B-Tier | 66 | 74 | 2026-06-13 |
| 8 | gohypergiant/agent-skills | 66 | Experimental | 66 | 62 | 2026-07-06 |
| 9 | VoltAgent/awesome-claude-code-subagents | 64 | B-Tier | 58 | 68 | 2026-06-24 |
| 10 | qdhenry/Claude-Command-Suite | 58 | B-Tier | 60 | 64 | 2026-03-01 |
Ranked by overall score: reliability 25%, skill leverage 20%, documentation 15%, maintenance 10%, safety 10%, evaluation readiness 10%, composability 5%, adoption capped at 5%. Full rubric on the methodology page. Full rubric.
Top entries in detail
A process-first skill framework for Claude Code: brainstorming, test-driven development, systematic debugging, and verification skills that…
Anthropic's first-party public skill collection: the de facto reference for SKILL.md structure, frontmatter conventions, and…
An opinionated 23-skill suite that staffs a solo developer's agent with the roles a startup would hire: CEO review, design review,…
Frequently asked questions
Why is adoption only 5% of the score?
Stars measure marketing reach, not reliability. Log-scaling plus the 5% cap means a popular-but-shallow repo cannot outrank a rigorous one. This is the methodology's core design decision.
Why are awesome-lists excluded?
Aggregators are discovery sources, not execution systems; they are indexed separately with their own scores but excluded from best-of rankings, which compare things you can actually install and run.