一键导入
sc-help
Show the catalog of installed Shadow Clone Kimi skills and what each one does
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。
菜单
Show the catalog of installed Shadow Clone Kimi skills and what each one does
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
Shadow Clone audit mode — produce an audit blueprint, then chain to the security checklist scan
Shadow Clone cleaner mode — repo organization + archival sweep. Evidence-gated, plan-first, fully reversible. Never deletes files.
Shadow Clone debug mode — systematic root-cause analysis and bug resolution with hardening
Shadow Clone docs mode — generate or refresh documentation for work currently in scope. Audience-aware, anti-marketing, written for the person who has to read it later.
Shadow Clone feature mode — build a new capability through dynamic waves, secure-by-design, production-ready
Shadow Clone optimize mode — measure-first performance improvement with clear baselines and targets
| name | sc-help |
| description | Show the catalog of installed Shadow Clone Kimi skills and what each one does |
| type | prompt |
| whenToUse | When the user asks what Shadow Clone skills or commands are available |
Show the user the following catalog exactly as written below, then ask if they want to use any of them. Do not invent skills; do not omit the "Status" column.
Every Shadow Clone surface is a skill: the umbrella /skill:sc, plus the
/skill:sc-<name> mode family. They all live as directories under
~/.kimi-code/skills/sc-*/SKILL.md once kimi/install.sh has run —
bash scripts/sc-doctor.sh verifies the install.
| Skill | What it does | Status |
|---|---|---|
/skill:sc | Activate Shadow Clone — walks the user through project init: detects existing setup, asks Type / Stack / Team / Stakes, derives a protocol shortlist, and writes (with user-gated overwrites) AGENTS.md, CLAUDE.md, docs/audit/ISSUE_TRACKER.md, and the .waves/ scaffold. Surfaces a branching migration plan if non-conforming branches exist. | ✅ Available |
/skill:sc-bootstrap | Verify the install is complete (bridge + protocols + reviewer persona + Kimi skills) and tell you the smallest next step to complete it. | ✅ Available |
/skill:sc-update | Check for and apply Shadow Clone updates — detects the install path and walks you through the right update route. | ✅ Available |
/skill:sc-help | Show this catalog. | ✅ Available |
/skill:sc-echo [claude|opencode|grok|kimi] | Enter paired-review mode — a second model reviews each completed work unit against AGENTS.md, returning VERDICT: APPROVE | REVISE | BLOCK | ERROR. Loop up to 3 rounds per unit. Default reviewer when Kimi builds: Claude (ask-claude.sh). | ✅ Available |
Each skill asks a short batch of context questions (3-5, via AskUserQuestion)
up front before starting Wave 0. Every mode adheres to the canonical engineering standards
deployed to ~/.claude/sc/protocols/ (see "Coding standards" below).
| Skill | What it does | Status |
|---|---|---|
/skill:sc-plan | Greenfield plan mode — produces MASTER_PLAN.md as a DAG of phases with explicit prerequisites, parallel branches, and load-bearing items. No timelines; work gets done when it gets done. | ✅ Available |
/skill:sc-sprint | Active-codebase sprint — one milestone, decomposed into a PR-sized task DAG (prerequisites, parallel-with, load-bearing flags). Produces SPRINT_PLAN.md with a pipeline diagram. No durations; phase-triggered check-ins. | ✅ Available |
/skill:sc-roadmap | Multi-milestone pipeline — Wave 0 strategic context, Wave 1 workstream DAG with prerequisites + parallelism + load-bearing flags, Wave 2 ROADMAP.md with decision gates. No timelines; the pipeline is shaped by the graph, not the calendar. | ✅ Available |
/skill:sc-feature | Feature mode — multi-wave implementation team for a focused capability. | ✅ Available |
/skill:sc-refactor | Refactor mode — safe restructure team with behavior-preserving discipline. | ✅ Available |
/skill:sc-debug | Debug mode — investigation team, root-cause analysis, hypothesis-then-test. | ✅ Available |
/skill:sc-optimize | Optimize mode — performance team, measure-first, micro-vs-macro tradeoff awareness. | ✅ Available |
/skill:sc-research | Research mode — open-ended investigation team for tech selection or spike work. | ✅ Available |
/skill:sc-audit | Audit mode — planning team that produces an audit blueprint, then chains into the security checklist scan. | ✅ Available |
/skill:sc-test-audit | Diagnose test-suite coverage gaps in an existing project — maps source surface against existing tests, flags missing integration tests where they make sense, surfaces security-sensitive paths without coverage. Read-only; does not run tests, does not generate them. | ✅ Available |
Smaller, single-purpose helpers for focused work that doesn't justify a full mode.
| Skill | What it does | Status |
|---|---|---|
/skill:sc-quick-fix | Targeted single-issue fix — minimal team, no waves, fastest path to a small, focused change. Gnosis-gated diagnosis, scope-guarded edits, verification per the user's chosen rigor. | ✅ Available |
/skill:sc-tests | Generate meaningful tests for the work currently in scope — surface-first contract discovery, framework-aware, no coverage theater. Cites Testing & Quality Assurance Protocol.md. | ✅ Available |
/skill:sc-docs | Generate documentation for the work currently in scope — audience-aware, anti-marketing voice, source-of-truth-first. Cites Documentation Standards for Software Teams.md. | ✅ Available |
/skill:sc-cleaner | Repo organization + archival sweep — evidence-gated, plan-first. Inventories stale/backup/misplaced files, writes CLEANUP_PLAN.md, and on approval moves them into archive/ (mirroring structure; git mv for tracked files, preserving history). Never deletes files (empty dirs only, gated); fully reversible. | ✅ Available |
These are bash scripts in scripts/ — run with bash scripts/<name>.sh:
| Script | What it does |
|---|---|
scripts/sc-doctor.sh | Verify the Shadow Clone install is healthy — checks every deployed path (including the Kimi skills), required CLIs, and the protocols deployment. |
scripts/sc-update.sh | Pull the latest release source, re-run the bridge installer, re-verify with sc-doctor. |
scripts/sc-last-verdict.sh | Print the verdict line from the most recent echo review in .sc/exchange/. |
The canonical engineering standards live in protocols/ in the repo and are
deployed to ~/.claude/sc/protocols/ by bridge/install.sh. Every /skill:sc-*
mode requires adherence to them; when a /skill:sc-echo review flags a protocol
violation, the finding cites the protocol filename.
Core standards (apply to every mode):
Functional Programming & Purity Protocol.mdComprehensive Code Quality and Consistency Protocol.mdSECURITY_CHECKLIST.mdError Handling & Resilience Protocol.mdAI-Assisted Development Protocol.mdGnosis Verification Protocol.md — load-bearing evidence gate: a bug that has not been verified is not a bugOperational protocols (govern HOW Shadow Clone runs):
Multi Agent Protocol.md — orchestrator-worker theory and concurrency capsShadow Clone Wave & Subagent Coordination Protocol.md — wave lifecycle,
role-to-clone mapping, team-size capping (hard cap 5 concurrent specialists),
Record Keeper contract, failure modes, audit loggingAdditional protocols (referenced by relevant modes):
Architecture & System Design Protocol.mdCode Efficiency & Performance Protocol.mdTesting & Quality Assurance Protocol.mdDocumentation Standards for Software Teams.mdAudit Protocol.mdDependency & Supply Chain Management Protocol.mdDevOps & Deployment Protocol.mdkimi/install.sh and ready to use.For the list of currently installed Kimi skills on this machine, look at
~/.kimi-code/skills/. Anything there is invokable as /skill:<directory-name>.