菜单
Scaffold a new Clade skill end-to-end — interviews for use cases and trigger phrases, generates SKILL.md + prompt.md with spec-validated frontmatter (bilingual triggers, NOT-for disambiguation), wires golden-set routing tests, and runs the lint gate before committing
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。
Clade goal-driven autonomous improvement loop (Blueprint architecture — deterministic pre/post phases + LLM supervisor/worker nodes, converges when goal met or max-iter hit). NOT the Claude Code built-in /loop (which polls a prompt on an interval like `/loop 5m /foo`) — if the user wants interval polling, route to the built-in.
Adaptive, resumable Git delivery across repositories and agent runtimes — probe policy, checkpoint coherent work, publish, review, integrate, and verify cleanup
Adaptive, resumable Git delivery across repositories and agent runtimes — probe policy, checkpoint coherent work, publish, review, integrate, and verify cleanup
Inspect or select a Clade execution connection while keeping agent runtime, inference provider, wire protocol, and model distinct. Use for Claude/Codex runtime selection, Anthropic/OpenAI/MiniMax/Kimi/custom gateways, model routing, or connection troubleshooting.
Inspect or select a Clade execution connection while keeping agent runtime, inference provider, wire protocol, and model distinct. Use for Claude/Codex runtime selection, Anthropic/OpenAI/MiniMax/Kimi/custom gateways, model routing, or connection troubleshooting.
Equipment manager for Claude Code projects — inventory local skills/agents/scripts, audit external upstream repos (red-flag + prompt-injection screening), and sync selectively after user approval. Absorbs container-layout skill repos AND single-skill-at-root repos (e.g. a company design-system repo). Project-agnostic.
| name | skill-new |
| description | Scaffold a new Clade skill end-to-end — interviews for use cases and trigger phrases, generates SKILL.md + prompt.md with spec-validated frontmatter (bilingual triggers, NOT-for disambiguation), wires golden-set routing tests, and runs the lint gate before committing |
| when_to_use | create a new skill, scaffold a skill, add a skill, write a new slash command, build a skill for X, 新建技能, 创建技能, 写个新技能 — NOT for hook generation (use /generate-hook), NOT for absorbing external skill repos (use /equip) |
| user_invocable | true |
| argument-hint | [skill-name or one-line description of what it should do] |
Scaffolds a new skill for the Clade skill registry (configs/skills/),
encoding the Agent Skills spec rules (agentskills.io) and Clade's own
conventions so every new skill is born consistent, discoverable, and tested.
SKILL.md (discovery layer) + prompt.md (execution body)
following the frontmatter hard rules enforced by validate-skills.pyorchestrator/tests/test_routing_eval.py so triggering is CI-guardedcommitter and reminds about ./install.sh deploy/skill-new # full interview
/skill-new pdf-extract # name known, interview the rest
/skill-new "summarize meeting notes into action items"
name = directory name, kebab-case, no "claude"/"anthropic"description ≤ 1024 chars, single line, WHAT + key capabilities, no < >when_to_use carries trigger phrases users would actually type (EN + 中文),
ending with NOT for X (use /y) clauses when siblings could confuse routingSKILL.md body under 500 lines (Agent Skills spec); execution detail lives
in prompt.md, bulky reference material in references/README.md inside the skill folder