Neeeophytee/finding-unknowns-skills

14 installable skills for Claude Code, OpenAI Codex, and Hermes: find unknowns, clarify requirements, manage context, test assumptions, and verify bug fixes with regression evidence.

What it solves

This project provides a collection of 14 installable "skills" (reusable instruction sets) designed to help coding agents identify and resolve "unknowns"—the gaps between a user's prompt and the actual codebase or real-world requirements—before they become expensive to fix during implementation.

How it works

The project uses the SKILL.md format, which is compatible with various AI coding agents (such as Claude Code, OpenAI Codex, Kimi Code CLI, and Cursor). These skills act as specialized commands or prompts that the agent can invoke to perform specific tasks like auditing context, brainstorming prototypes, or conducting "blindspot passes" to surface unknown unknowns.

Who it’s for

Developers using AI coding agents who want to move from simple prompting to a more structured workflow for clarifying requirements, testing assumptions, and auditing the agent's own context window.

Highlights

  • Task-Level Skills: Includes tools for the "Before", "During", and "After" phases of development (e.g., blindspot-pass, implementation-plan, change-quiz).
  • Context-Level Skills: Tools to optimize the agent's system instructions and context window (e.g., context-audit, progressive-disclosure).
  • Context Engineering: Implements techniques to reduce context window bloat and avoid contradictions in agent instructions.
  • Broad Compatibility: Supports installation via npx skills add or as plugins for multiple major AI coding agents.

Related

  • Project
    MengTo/Skills

    A curated library of procedural playbooks and prompts that turn complex design and coding workflows into reusable skills for AI agents to ensure consistent, high-quality output.

  • Project
    xstongxue/best-skills

    A collection of high-quality SKILL.md files for AI agents (Cursor, Claude Code, etc.) that enable automatic task triggering for academic writing, diagramming, and development workflows.

  • Project
    addyosmani/agent-skills

    A library of production-grade engineering workflows and quality gates for AI coding agents to ensure they follow senior-level software development best practices.

  • Project
    staruhub/ClaudeSkills

    A collection of reusable agent skills in the SKILL.md format that transforms complex professional workflows into installable packages for AI agents.

  • Project
    mxyhi/ok-skills

    OK Skills is a library of 31 ready‑to‑use *skill packs* for AI coding agents (Codex, Claude Code, Cursor, etc.). Each skill lives in its own directory with a `SKILL.md` that tells the agent how to run the underlying script (e.g., browser automation, docs lookup, planning, design). Install by cloning into `~/.agents/skills` (or the Autohand equivalent), copy the `CLAUDE_AGENTS.md` sections into your agent’s config, and then ask the agent to “Use <skill‑name> …”. The repo provides categories such as Research, Planning, Automation, Frontend, and Utilities, and serves developers who want portable, repeatable AI‑agent capabilities instead of ad‑hoc prompts.