菜单
Multi-skill workflow composer. Activates ONLY for tasks that require combining multiple skills in sequence or resolving skill conflicts. Single-skill tasks: return [] — individual skills self-trigger via their own descriptions.
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
Visualize an agent's perceive→think→act loop. Inspects tool calls, latency per step, decision branches, retry patterns. Surfaces bottlenecks. Triggers on "trace agent", "why is agent slow", "agent loop analysis", "what did the agent do".
Compute context window budget for a prompt/agent — sums system prompt, tool descriptions, examples, conversation history; flags overruns; suggests cuts. Triggers on "context budget", "how much context", "will this fit", "budget my prompt".
Treat dataset / eval JSONL as code — lint, validate schema, detect duplicates, leakage between train/eval, label imbalance, PII. Karpathy "Software 2.0" lens. Triggers on "audit dataset", "lint jsonl", "check for leakage", "validate eval data".
Enforce "evals over vibes" — before any prompt, agent, or model change, require an eval set with success criteria and baseline measurement. Blocks vibe-only changes. Triggers on "tune prompt", "improve agent", "change model" without prior eval set.
Build a 10-30 case eval dataset from a PRD, skill, or prompt — covering golden path, edge cases, and adversarial inputs. Output as JSONL with input/expected/category. Triggers on "build eval set", "create evals", "generate test cases for prompt".
Break down any ML/AI concept to first principles — strip jargon, derive from basics, build physical/mathematical analogy. Karpathy intuition-building style. Triggers on "first principles", "explain like I'm 5", "derive from scratch", "why does X actually work".
| name | skills-selector |
| description | Multi-skill workflow composer. Activates ONLY for tasks that require combining multiple skills in sequence or resolving skill conflicts. Single-skill tasks: return [] — individual skills self-trigger via their own descriptions. |
Scope: Compose multi-skill workflows. NOT a replacement for individual skill auto-triggers.
[] (most common — correct behavior)Return [] when:
smart-dispatch alone handles itSingle-skill requests where individual auto-trigger is correct:
| User intent | Self-triggers |
|---|---|
| "commit", "cn", "save" | semantic-commit / git-commit |
| "review code", "audit" | code-review / caveman-review |
| "debug", "bug", "fix error" | debug-workflow / diagnose |
| "build UI", "design component" | frontend-design / ui-ux-pro-max |
| "write post", "copy", "blog" | copywriting / lemon-blog-post |
| "create PR" | git-pr-create |
| "write tests" | tdd |
| "build MCP tool" | mcp-builder |
| "create agent" | Agent Development |
Activate ONLY when the task is explicitly multi-phase or combines domains:
| Pattern | Phase 1 | Phase 2 | Phase 3 |
|---|---|---|---|
| "find bug and commit the fix" | debug-workflow | semantic-commit | — |
| "fix bug and open PR" | debug-workflow | git-pr-create | — |
| "review and create PR" | code-review | git-pr-create | — |
| "implement feature and write tests" | smart-dispatch | tdd | — |
| "implement feature, test, and commit" | smart-dispatch | tdd | semantic-commit |
| "design UI and implement" | frontend-design | smart-dispatch | — |
| "build MCP and write docs" | mcp-builder | copywriting | — |
| "write post and publish" | lemon-blog-post | (publish flow) | — |
| "triage issues and plan" | triage | to-prd | — |
| "refactor and review" | smart-dispatch | code-review | — |
| "audit architecture and improve" | improve-codebase-architecture | smart-dispatch | — |
| "create changelog and PR" | git-changelog | git-pr-create | — |
| "diagnose perf and fix" | diagnose | smart-dispatch | — |
| "write skill and test" | write-a-skill | manual test | — |
| "prototype then build" | prototype | smart-dispatch | — |
| "spec then implement" | to-prd | smart-dispatch | — |
| "debug mobile release cert" | mobile-fix-release-certificates | debug-workflow | — |
| "create social + blog for feature" | social-content | lemon-blog-post | — |
debug-workflow + code-review simultaneously — pick one; load second after first completes.frontend-design + ui-ux-pro-max together — ui-ux-pro-max supersedes.diagnose + debug-workflow together — diagnose is for hard/perf bugs; debug-workflow for standard bugs.smart-dispatch is always available as base layer; do not count it as a "skill slot."SKILLS PLAN:
Phase 1: <skill-name> — <why>
Phase 2: <skill-name> — <trigger condition>
Phase 3: (optional)
CONFLICT CHECK: <none | resolved: X supersedes Y>
If no multi-skill workflow detected:
SKILLS PLAN: []
Reason: single-skill task — <skill-name> self-triggers.