Personal Agent Skills catalog for Claude Code and Codex.
This repository is designed to work with the skills CLI:
npx yskill
npx yskill --list
npx yskill --preset init
npx yskill --preset frontend
npx yskill --skill designyskill is a small npm shortcut around:
npx skills add yikZero/skills --list
npx skills add yikZero/skills --skill designWith no arguments, npx yskill opens a menu:
What do you want to install?
> Project init
Core recommended
Frontend bundle
Pick individual skills
Install all
In non-interactive shells, pass a selection explicitly so it cannot accidentally install everything:
npx yskill --preset init
npx yskill --preset default
npx yskill --preset frontend
npx yskill --skill find-docs
npx yskill --allAdd agent flags only when you want to target a specific app, for example
-a codex, -a claude-code, or both. Add -y only for non-interactive
installs.
For local development before publishing:
npx skills add /path/to/skills --list
npx skills add /path/to/skills --skill designskills/
<skill-name>/
README.md
SKILL.md
references/
scripts/
assets/
docs/
ADDING_SKILLS.md
templates/
skill/
scripts/
validate-skills.mjs
skills/ is the only installable catalog. Each skill has a short README.md for GitHub browsing and install commands; SKILL.md remains the agent entrypoint. Put drafts in skills/.experimental/ only when they are valid but intentionally hidden from normal installs by metadata.internal: true.
design: focused frontend UI review and refinement workflow adapted frompbakaus/impeccable, exposed asaudit,critique,polish,harden,layout,typeset,colorize,clarify, anddistill.chat-sdk: build and review Vercel Chat SDK bots across Slack, GitHub, Discord, Teams, and other chat platforms using SDK-native adapters, events, streaming, state, cards, and webhooks.find-docs: Context7 CLI workflow for fetching current library, framework, SDK, API, and CLI documentation.project-bootstrap: turn a new product, app, automation, or library idea into a ready-to-build project baseline with current-docs discovery, runtime or framework selection, control docs, repo-local helper skills, validation, and a first-workflow handoff.postgres-best-practices: design, query, migrate, and operate PostgreSQL with production-proven rules for schema and data types, indexing, query tuning, zero-downtime migrations, and scaling, synthesized from OpenAI's scaling playbook, the PostgreSQL wiki, and strong_migrations.karpathy-guidelines: behavioral guardrails inspired by Andrej Karpathy's observations for reducing overcomplication, broad edits, hidden assumptions, and unverifiable coding work.react-composition: review and refactor React component APIs, state ownership, and composition patterns.react-performance: review and improve React or Next.js performance using Vercel-style rules.shadcn-ui: add, customize, and review shadcn/ui components with CLI-backed workflows.skill-authoring: create, update, and validate portable Agent Skills.tailwind-design-system: build and audit Tailwind CSS design systems, tokens, variants, and component patterns.web-ui-audit: run terse file:line audits against web UI implementation guidelines.workflow: install, configure, build, debug, and self-host Workflow SDK workflows, prioritizing TanStack Start, Next.js, and Vite while still linking to broader docs and examples.pi-coding-agent-sdk: source-backed workflows and recipes for embedding Pi Coding Agent SDK sessions, runtime flows, custom tools, custom skills, resource loading, streaming events, and test harnesses in TypeScript host apps.
Project installs:
- Claude Code:
.claude/skills/<skill-name>/ - Codex:
.agents/skills/<skill-name>/
Global installs:
- Claude Code:
~/.claude/skills/<skill-name>/ - Codex:
~/.codex/skills/<skill-name>/
The default skills CLI install mode uses a canonical copy plus symlinks where supported, which keeps multi-agent installs easier to update. In project installs, .agents/skills/ is the canonical location; Claude Code gets a .claude/skills/<skill-name> symlink when the target project already has a .claude/ directory. Use --copy only when symlinks are not appropriate.
npm run validate
npm run test:cli
npx skills@latest add . --list
npx skills@latest add . --skill design
npx skills@latest update design -p -ynpx skills add yikZero/skills --list temporarily clones the remote repository to inspect skills; it should not write .agents/, .claude/, or skills-lock.json into this source repository.
Run real remote installs from a consuming project, not this source repository:
npx yskill
npx yskill --preset init
npx yskill --preset default
npx yskill --preset frontend
npx yskill --skill design
npx skills@latest add yikZero/skills@designSee docs/SKILLS_CLI_NOTES.md for source formats, lock behavior, update rules, and safe local testing.
Use docs/ADDING_SKILLS.md as the process checklist. Keep AGENTS.md as the source of truth for agent behavior in this repository.
That doc also includes copy-paste prompts for asking an AI agent to add a new skill or migrate an existing skill safely.