Skill Index — Progressive Disclosure Engine
TL;DR
- Use when searching for a CodyMaster or community skill
- Layered: L1 index (always loaded) → L2 summary → L3 full
- Includes: discovery, search, mastery (consolidated in v6)
Save 90%+ tokens. Agents scan the index first, load full skills only when executing. Inspired by Loki Mode's 3-Layer Progressive Disclosure memory architecture.
The Problem
Traditional approach: Load every SKILL.md to decide which skill to use.
- 30+ skills × ~3000 tokens each = 90,000+ tokens just for discovery
- Agent only needs ~500 tokens to make the choice
- 89,500 tokens wasted on reading full skills
The Solution: 3-Layer Loading
┌────────────────────────────────────────────────┐
│ LAYER 1: INDEX (~100 tokens per skill) │
│ Always loaded. Quick scan. "What exists?" │
│ Name + Domain + Triggers + 1-line summary │
└─────────────────────┬──────────────────────────┘
│ (match found?)
v
┌────────────────────────────────────────────────┐
│ LAYER 2: SUMMARY (~300 tokens per skill) │
│ Load when choosing. "Is this the right one?" │
│ Description + When to use + Integration table │
└─────────────────────┬──────────────────────────┘
│ (confirmed?)
v
┌────────────────────────────────────────────────┐
│ LAYER 3: FULL SKILL.md (1500-5000 tokens) │
│ Load ONLY during execution. "How to use it." │
│ Complete instructions, examples, red flags │
└────────────────────────────────────────────────┘
Layer 1: Skill Index
Always loaded by agents at session start. Maximum 100 tokens per skill.
Engineering Swarm 🔧
| Skill | Triggers | Summary |
|---|---|---|
cm-tdd | test, TDD, red-green-refactor | Red-Green-Refactor cycle before any implementation |
cm-debugging | bug, error, fix, debug, broken | Root cause analysis before fixing. 5-phase investigation |
cm-quality-gate | deploy, ship, verify, quality, gate | 6-gate verification: static analysis → blind review → ship |
cm-test-gate | test setup, CI, test infrastructure | Setup 4-layer test gate for any project type |
cm-code-review | review, PR, feedback, branch | Request reviews, handle feedback, complete branches |
cm-codeintell | understand codebase, call graph, impact, architecture, what calls X, what breaks | AST knowledge graph + architecture diagrams + smart context (30% fewer tokens) |
cm-reactor | rewrite, pivot, migrate, direction change, architecture wrong, restructure | TRIZ-powered strategic codebase re-direction — 5-phase incremental migration |
cm-clean-code | clean, dead code, refactor, code smell, DRY, SOLID, naming, messy | 7-point hygiene gate: dead code, imports, naming, SRP, DRY, nesting |
Operations Swarm ⚙️
| Skill | Triggers | Summary |
|---|---|---|
cm-safe-deploy | deploy, staging, production, release | Multi-gate deploy pipeline with rollback strategy |
cm-identity-guard | git push, deploy, Cloudflare, Supabase, account | Verify identity before any push/deploy to prevent wrong-account |
cm-git-worktrees | branch, isolate, worktree, parallel | Isolated git worktrees for feature work |
cm-terminal | command, terminal, run, execute | Safe terminal execution with progress logging |
Security Swarm 🔒
| Skill | Triggers | Summary |
|---|---|---|
cm-secret-shield | secret, token, leak, security, scan, pre-commit, gitleaks, rotate | Defense-in-depth: pre-commit hooks, repo scanning, token lifecycle |
Product Swarm 🎨
| Skill | Triggers | Summary |
|---|---|---|
cm-brainstorm-idea | analyze, enhance, improve, initiative, evaluate, review product | Strategic analysis gate: 9 Windows + Double Diamond → 2-3 qualified options |
cm-planning | plan, design, brainstorm, feature | Brainstorm intent → write implementation plan → then code |
cm-ux-master | UI, UX, design, interface, usability | 48 UX Laws + 37 Design Tests + Figma/Stitch integration |
cm-ui-preview | preview, visual, mockup, render UI, wireframe | Live UI concept generation via Google Stitch or Pencil MCP |
cm-dockit | docs, documentation, knowledge base, SOP | Generate complete knowledge base from codebase |
cm-project-bootstrap | new project, init, bootstrap, setup | Full project setup: design system → staging → CI → deploy |
cm-jtbd | customer discovery, JTBD, jobs to be done, product-market fit, why users churn | JTBD canvas: Switch Interview → Outcome Metrics → Opportunity Scoring |
Growth Swarm 📈
| Skill | Triggers | Summary |
|---|---|---|
cm-content-factory | content, blog, article, marketing | AI content engine: research → generate → audit → deploy |
cm-ads-tracker | pixel, tracking, GTM, Meta, TikTok, Google Ads | Complete conversion tracking setup across platforms |
cm-cro-methodology | conversion, A/B test, landing page, funnel | CRO audit: funnel mapping → persuasion → objection handling |
Orchestration Swarm 🎯
| Skill | Triggers | Summary |
|---|---|---|
cm-execution | execute, implement, plan, RARV, batch | Execute plans: batch mode, subagent-driven, parallel, or RARV |
cm-continuity | memory, context, CONTINUITY, learnings | Working memory protocol: read at start, update at end |
cm-skill-mastery | which skill, skill list, discover, help | Meta-skill: when to invoke skills, how to create new ones |
cm-skill-health | skill broken, skill drift, missing skill, stale skill, health | Audit a skill using docs drift, retro notes, validation, and gates |
cm-skill-evolution | repair skill, evolve skill, fix skill, derive skill | Recover a degraded skill with FIX, DERIVED, or CAPTURED mode |
cm-skill-search | find skill, lookup skill, which skill, search skill | Find the right skill through indexes, suggestions, and repo search |
cm-skill-share | export skill, import skill, share skill, move skill | Package and transfer skill folders safely across repos and machines |
cm-safe-i18n | translate, i18n, language, localize | Safe translation with multi-pass batching and audit gates |
cm-skill-chain | chain, pipeline, workflow, multi-step, full process | Compose skills into automated multi-step pipelines |
cm-deep-search | semantic search, find docs, large codebase, qmd | Optional power-up for semantic search across large projects |
cm-readit | audio, TTS, read aloud, voice, speech, SpeechSynthesis, MP3 player | Web audio engine — TTS reader, pre-recorded MP3 player, Voice CRO triggers |
cm-how-it-work | how does X work, explain, architecture overview | Explain how a system, feature, or codebase works |
Workflow Commands 🖥️
| Skill | Triggers | Summary |
|---|---|---|
cm-start | /cm-start, start workflow, begin objective, new task | Orchestrate full cm-* workflow from objective to production code |
cm-dashboard | /cm-dashboard, kanban, task board, show tasks, status board | Render Kanban board from OpenSpec tasks.md and cm-tasks.json — visual task overview |
cm-status | /cm-status, quick status, progress, what's next, blocked | Ultra-concise progress summary: done / next / blocked |
Planned (not yet implemented) 🧪
Skills below do not have SKILL.md files yet.
| Skill | Triggers | Summary |
|---|---|---|
mom-test (planned) | customer interview, validate idea, leading questions | Customer interviews without leading questions |
release-it (planned) | production, circuit breaker, timeout, chaos | Build production-ready systems with stability patterns |
medical-research (planned) | medical, OB/GYN, clinical, evidence-based | Evidence-based medical writing with citation standards |
tailwind-mastery (planned) | Tailwind, utility-first, responsive, v4 | Tailwind CSS u |