菜单
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.
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.
Integrate an exact reviewed PR under repository policy, choose truthful history semantics, and verify cleanup
| name | equip |
| description | 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. |
| when_to_use | equip, upstream, absorb skill, absorb a design-system repo, update skill from github, review external skill, sync skill, inventory skills, 装备, 更新外部技能, 审核上游, 管理插件来源 — NOT for first-party feature development (just write code) |
| argument-hint | <command> [args] # inventory | audit <repo|id> | sync <id> [--apply] | diff <id> | list | add <repo> | remove <id> |
| user_invocable | true |
A project-agnostic skill for managing "absorbed" assets (skills, agents, scripts) that originated in external GitHub repos. It treats each upstream as a supplier whose shipments must be reviewed before adoption, not blindly synced.
| Command | Purpose |
|---|---|
/equip inventory | Scan current project; classify every asset as native / absorbed / modified-absorbed / orphan; write .claude/equipment/inventory.yaml |
/equip audit <repo|id> | Clone upstream, apply red-flag checks to each skill, and bind the decision report to the exact audited commit |
/equip sync <id> [--apply] | Refresh upstream, verify its exact commit matches the audit, then perform a 3-way merge; dry-run by default |
/equip diff <id> | Show per-file delta between registered last_synced_commit and remote HEAD |
/equip list | Show registered upstreams with ahead/behind status |
/equip add <repo> | Register a new upstream (interactive: select which local paths it covers) |
/equip remove <id> | Unregister an upstream (keeps files, removes tracking) |
/equip inventory # first step for any project
/equip audit AgriciDaniel/claude-seo # full evaluation
/equip sync claude-seo # dry-run: show what would change
/equip sync claude-seo --apply # actually write changes
/equip diff claude-seo # inspect drift
/equip audit . # self-audit current project as if upstream
--apply fails closed if the fetched commit differs from the exact commit in the audit report; re-run /equip audit to approve drift/equip picks.claude/equipment/audits/<id>-<date>.md that the user can read and edit