Skip to content

Agent skills

Agent skills are folders of instructions and scripts that teach Claude Code, Codex, Cursor and other agents a task. Each one lists what it can run and how to install it.

Showing 2,041–2,070 of 2,929

  1. Fusing scores from multiple searches into a single ranked result (RRF, DBSF, custom fusion)

    • Open source
    • 269 GitHub stars
    • Knowledge and memory
  2. Constructing prefetch queries for hybrid retrieval, including BM25/dense and multi-field setups, and choosing a lexical retrieval method

    • Open source
    • 269 GitHub stars
    • Knowledge and memory
  3. Expanding the candidate pool via relevance feedback, as an alternative to reranking when a dense retriever is too weak

    • Open source
    • 269 GitHub stars
    • Knowledge and memory
  4. Probe thinking through iterative rounds of useful questions that expose assumptions and blind spots.

    • Open source
    • 120 GitHub stars
  5. Make an abstract idea concrete through mechanisms, examples, or situations.

    • Open source
    • 120 GitHub stars
  6. Identify candidate next steps for continuing the current thinking.

    • Open source
    • 120 GitHub stars
  7. Identify the important unresolved parts of our thinking.

    • Open source
    • 120 GitHub stars
  8. Derive observable predictions from the current understanding.

    • Open source
    • 120 GitHub stars
  9. Prioritize the current questions or branches by expected value for the goal.

    • Open source
    • 120 GitHub stars
  10. Learn a concept by reconstructing why it had to be invented.

    • Open source
    • 120 GitHub stars
  11. Explore a different representation or perspective on the current problem.

    • Open source
    • 120 GitHub stars
  12. Recover the current state of our thinking.

    • Open source
    • 120 GitHub stars
  13. Synthesize the important ideas in the current thinking into one coherent model.

    • Open source
    • 120 GitHub stars
  14. Reconstruct how our thinking evolved through meaningful phases.

    • Open source
    • 120 GitHub stars
  15. Transfer the current insight to other problems, domains, or situations.

    • Open source
    • 120 GitHub stars
  16. Map what is known, unknown, implicit, or potentially missing about a problem.

    • Open source
    • 120 GitHub stars
  17. Verify the important claims in the current thinking against the available evidence.

    • Open source
    • 120 GitHub stars
  18. Guide for using the dhub CLI — the AI skill manager for data science agents

    • Open source
    • 116 GitHub stars
  19. Distill a top-level principle into knowledge/rules.md

    • Open source
    • 107 GitHub stars
  20. Enforces coding best practices: deep-read before modify, LSP-first navigation, TDD red-green-refactor, minimal changes, self-review, verific

    • Open source
    • 107 GitHub stars
  21. Fetch current library/framework documentation via Context7

    • Open source
    • 107 GitHub stars
    • Databases
  22. Systematic debugging: reproduce → hypothesize → verify → fix

    • Open source
    • 107 GitHub stars
  23. Branch completion workflow: verify tests → present 4 options (merge/PR/keep/discard) → execute → cleanup worktree

    • Open source
    • 107 GitHub stars
  24. Capture knowledge into episodes.md

    • Open source
    • 107 GitHub stars
  25. Generate professional PDF documents with correct CJK (Chinese/Japanese/Korean) rendering

    • Open source
    • 107 GitHub stars
    • Files and documents
  26. Full plan lifecycle: deep understanding → write plan with TDD checklist → parallel review → Ralph Loop execution

    • Open source
    • 107 GitHub stars
  27. Code and plan review with multi-angle dispatch

    • Open source
    • 107 GitHub stars
  28. Agent self-learning: promote recurring episodes to rules, capture complex corrections

    • Open source
    • 107 GitHub stars
  29. Create high-quality, production-ready skills from scratch

    • Open source
    • 107 GitHub stars