菜单
Scaffold new AI agent skills with proper symlinking across all IDE tools (Cursor, Claude Code, Codex, etc.). Use when creating a new skill, adding a local skill, or when the user wants to write a custom skill -- either workspace-wide or for a specific project.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Principle-engineering posture for production-grade code: reads the repo first, plans before code, matches conventions, pulls latest docs over training recall, and ships the simplest correct change that holds the bar — proper algorithms and data structures, idempotent writes, schema+queries+indexes as one artefact, typed errors, tests in the same diff. Substrate-agnostic; defers to peer skills on their lanes. Use for non-trivial planning, design, implementation, review, or refactoring; RCA and debugging; performance and optimization work; changes touching a database schema, security, infrastructure, or a public API; hardening inherited, vibe-coded, or LLM-generated code (dependency/CVE and migration audits); and over-engineering cleanup ("simplest solution," "YAGNI," "what can we delete").
Runs a doer -> verifier-panel -> consensus loop to verify a deliverable before it ships. An orchestrator freezes acceptance criteria before implementation, dispatches a doer, then convenes a context-walled panel of independent verifiers - including an adversary with an explicit must-oppose mandate - for evidence-anchored review adjudicated to a SHIP / SHIP_WITH_CAVEATS / ITERATE / BLOCK / ESCALATE verdict logged to a ledger. Use for multi-agent verification of any artifact - code slices, plans, documents, audits - whenever asked to verify a deliverable, vet a plan, run a consensus review or independent review, set up a doer-verifier loop, or gate a ship decision. Works on any platform with parallel subagents; degrades to sequential fresh-context sessions without them. Not for trivial single-file edits or ordinary code review.
| name | scaffold-skill |
| description | Scaffold new AI agent skills with proper symlinking across all IDE tools (Cursor, Claude Code, Codex, etc.). Use when creating a new skill, adding a local skill, or when the user wants to write a custom skill -- either workspace-wide or for a specific project. |
Creates new skills using the workspace script, which handles name validation, directory placement, and symlinking to all AI tool directories.
Ask the user if unclear.
root-config/.agents/skills/, symlinked to the
workspace root so all projects and AI tools (.agents/skills/, .claude/skills/,
skills/) can discover it.<repo>/.agents/skills/, symlinked into that
project's AI tool directories only.Run from the workspace/ directory:
# Workspace-wide
npm run skills:create -- --name <skill-name>
# Project-specific
npm run skills:create -- --name <skill-name> --project <repo>
This validates the name, creates the directory with a template SKILL.md, and
runs setup to symlink the skill into .claude/skills/ and
skills/ so every AI tool can discover it.
Naming rules: lowercase letters, numbers, hyphens, underscores, and dots. Max 64 chars. Must start and end with a letter or number.
Do not write the SKILL.md content directly. Delegate to your IDE's skill-authoring guide for quality content and proper requirements gathering.
Search your available skills for one named create-skill. Common locations:
~/.cursor/skills-cursor/create-skill/SKILL.md~/.claude/skills-cursor/create-skill/SKILL.mdIf found: Read it and follow its complete workflow from the start --
including any discovery or requirements-gathering phase (asking the user
questions about purpose, triggers, format, etc.). Use the scaffolded
SKILL.md as your starting point; the guide tells you how to fill it in.
If not found: Edit the generated SKILL.md directly:
description with a specific, third-person
description that includes both WHAT the skill does and WHEN to use it.Confirm the skill appears in the expected locations:
npm run skills:list
If you cannot run the script, create the skill directory and SKILL.md manually
under the correct .agents/skills/ path. Then ask the user to run:
cd workspace && npm run skills:setup
This re-syncs all symlinks.