[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-92446":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":9,"languages":9,"totalLinesOfCode":9,"stars":10,"forks":11,"watchers":12,"openIssues":13,"contributorsCount":13,"subscribersCount":13,"size":13,"stars1d":13,"stars7d":13,"stars30d":14,"stars90d":13,"forks30d":13,"starsTrendScore":13,"compositeScore":15,"rankGlobal":9,"rankLanguage":9,"license":16,"archived":17,"fork":17,"defaultBranch":18,"hasWiki":19,"hasPages":17,"topics":20,"createdAt":9,"pushedAt":9,"updatedAt":27,"readmeContent":28,"aiSummary":29,"trendingCount":13,"starSnapshotCount":13,"syncStatus":30,"lastSyncTime":31,"discoverSource":32},92446,"finding-unknowns-skills","Neeeophytee\u002Ffinding-unknowns-skills","Neeeophytee","8 installable skills for Claude Code and OpenAI Codex, for finding your unknowns before they get expensive: blindspot pass, brainstorms, interviews, references, plans, notes, pitches, and a merge quiz. Community distillation, not official Anthropic.",null,163,17,108,0,39,47.67,"Other",false,"main",true,[21,22,23,24,25,26],"agent-skills","ai-agents","claude","claude-code","prompting","skills","2026-07-22 04:02:06","# Finding-Unknowns Skills\n\n**8 installable skills that make Claude help you find what you don't know — before it gets expensive to fix.**\n\nThe map is not the territory. Your prompt is a map; the codebase and the real world are the territory. The gap between them is your *unknowns*, and with strong models the quality of the work is bottlenecked by how well you clarify them. These skills turn that idea, from [Thariq Shihipar's](https:\u002F\u002Fthariqs.github.io\u002Fhtml-effectiveness\u002Funknowns\u002F) essay *A Field Guide to Fable: Finding Your Unknowns*, into commands you can run in Claude Code, OpenAI Codex, or any agent that reads the [agentskills.io](https:\u002F\u002Fagentskills.io) SKILL.md format.\n\n> Community project. Distilled, with attribution, from a public essay by Thariq Shihipar (Anthropic, Claude Code team). **Not an official Anthropic repository.**\n\n## The idea in one table\n\n|  | **Known** | **Unknown** |\n|---|---|---|\n| **Known** | What's in your prompt | What you know you haven't figured out |\n| **Unknown** | So obvious you'd never write it down, but you'd recognize it on sight | What you haven't considered at all |\n\nEvery skill below is a cheap way to move something out of the bottom row before implementation makes it expensive.\n\n## The skills\n\n| Skill | Phase | One line |\n|---|---|---|\n| [`blindspot-pass`](skills\u002Fblindspot-pass\u002FSKILL.md) | Before | Surface your unknown unknowns in an unfamiliar area, then help you prompt better |\n| [`brainstorm-prototypes`](skills\u002Fbrainstorm-prototypes\u002FSKILL.md) | Before | Throwaway variations you can react to, for taste you can't verbalize |\n| [`interview-me`](skills\u002Finterview-me\u002FSKILL.md) | Before | One question at a time, architecture-changing questions first |\n| [`reference-hunt`](skills\u002Freference-hunt\u002FSKILL.md) | Before | Use working source code as the spec, even across languages |\n| [`implementation-plan`](skills\u002Fimplementation-plan\u002FSKILL.md) | Before | A plan that leads with the decisions you're most likely to change |\n| [`implementation-notes`](skills\u002Fimplementation-notes\u002FSKILL.md) | During | Log every deviation from the plan so the next attempt learns from this one |\n| [`pitch-packager`](skills\u002Fpitch-packager\u002FSKILL.md) | After | Bundle spec + prototype + notes into a buy-in doc for reviewers |\n| [`change-quiz`](skills\u002Fchange-quiz\u002FSKILL.md) | After | A comprehension quiz you must pass before you merge |\n\n## Install\n\n**One command, any agent (recommended):** Vercel's `skills` CLI auto-detects your coding agent (Claude Code, Cursor, Codex, Copilot, Gemini, and more) and installs the skills into the right place for each:\n\n```\nnpx skills add Neeeophytee\u002Ffinding-unknowns-skills\n```\n\nAdd `--list` to preview the 8 skills first, or `--skill blindspot-pass` to install just one. (Discoverable on [skills.sh](https:\u002F\u002Fskills.sh).)\n\n**As a Claude Code plugin (all 8 skills):**\n\n```\n\u002Fplugin marketplace add Neeeophytee\u002Ffinding-unknowns-skills\n\u002Fplugin install finding-unknowns@finding-unknowns-skills\n```\n\n**Manually (pick the skills you want):** copy any `skills\u002F\u003Cname>\u002F` folder into your project's `.claude\u002Fskills\u002F` directory (or `~\u002F.claude\u002Fskills\u002F` for all projects).\n\n**The one-file version:** if you'd rather have the whole approach as passive guidance instead of commands, drop [`CLAUDE.md`](CLAUDE.md) (Claude Code) or [`AGENTS.md`](AGENTS.md) (Codex and other AGENTS.md-reading agents) into your project root, or append it to your existing one.\n\n### Use in Cursor\n\nThe `npx skills add Neeeophytee\u002Ffinding-unknowns-skills` command above detects Cursor and installs the skills into `~\u002F.cursor\u002Fskills\u002F` for you — no manual step. (Verified: Vercel's CLI discovers all 8 skills in this repo.)\n\n### Use in OpenAI Codex\n\nThe skills use the same `SKILL.md` format Codex reads natively, so no conversion is needed. Either run `npx skills add Neeeophytee\u002Ffinding-unknowns-skills` (it detects Codex), or copy them into a Codex skill location:\n\n```\ngit clone https:\u002F\u002Fgithub.com\u002FNeeeophytee\u002Ffinding-unknowns-skills\ncp -r finding-unknowns-skills\u002Fskills\u002F* ~\u002F.agents\u002Fskills\u002F        # all projects\n# or, per project:  cp -r finding-unknowns-skills\u002Fskills\u002F* your-repo\u002F.agents\u002Fskills\u002F\n```\n\nCodex detects skill changes automatically. For the passive-guidance version, drop [`AGENTS.md`](AGENTS.md) into your project root — Codex reads it before doing any work. Tested with Codex CLI v0.143: all 8 skills and their trigger descriptions load into the model-visible prompt (verifiable yourself with `codex debug prompt-input`). (Paths per the [Codex skills docs](https:\u002F\u002Fdevelopers.openai.com\u002Fcodex\u002Fskills).)\n\n## When to reach for which\n\n- New to a part of the codebase, or a whole domain → `blindspot-pass`\n- You'll know it when you see it (design, UX, tone) → `brainstorm-prototypes`\n- You've brainstormed but ambiguity remains → `interview-me`\n- You can't describe it, but some code somewhere does it right → `reference-hunt`\n- Ready to build → `implementation-plan`, then keep `implementation-notes` running\n- Built → `pitch-packager` for buy-in, `change-quiz` before you merge\n\nSee [EXAMPLES.md](EXAMPLES.md) for real prompts.\n\n## Credit\n\nThe techniques and the unknowns framing come from Thariq Shihipar's essay and his [companion artifacts](https:\u002F\u002Fthariqs.github.io\u002Fhtml-effectiveness\u002Funknowns\u002F). This repo distills them into the SKILL.md format with original instruction text; read the essay for the full reasoning and the Fable 5 launch-video story that motivates it. Coverage: [The Decoder](https:\u002F\u002Fthe-decoder.com\u002Fanthropic-developer-shares-prompting-tips-for-fable-5-that-focus-on-finding-your-own-blind-spots-first\u002F).\n\n## License\n\n[MIT](LICENSE) for the skill text in this repo. The underlying techniques belong to their author; attribution above.\n\n---\n\n\u003Csub>Maintained alongside [awesome-ai-workflows](https:\u002F\u002Fgithub.com\u002FNeeeophytee\u002Fawesome-ai-workflows), a list of AI workflows re-checked by CI, and [FlowStacks](https:\u002F\u002Fflowstacks.xyz), where each recipe carries a machine-verified badge. If this repo is useful, a star helps the next person find it.\u003C\u002Fsub>\n","这是一个面向AI编程助手（如Claude Code、OpenAI Codex）的实用技能集，提供8个可安装的提示工程工具，帮助开发者在编码前识别知识盲区、澄清隐含假设、验证设计决策，并在开发中记录偏差、交付后巩固理解。核心特点包括基于‘未知分类’理论（Known\u002FUnknown二维框架）设计的结构化技能流程，支持标准化SKILL.md格式，兼容多平台代理，安装便捷。适用于软件开发早期需求澄清、技术方案评审、团队知识对齐及代码审查准备等场景，尤其适合中大型项目中降低返工成本与沟通熵增。",2,"2026-07-09 02:30:02","CREATED_QUERY"]