Skip to content

Skills Library

Quick Reference

  • Total Skills: 30+ across 5 swarms
  • Format: Universal — works on any AI coding platform
  • Open Source: Full SKILL.md content visible for every skill
  • License: Open and transparent

Cody Master skills are the instructions that turn AI agents into disciplined engineers. Each skill is a markdown file (SKILL.md) with structured protocols that any AI agent can follow. Browse them all below — nothing is hidden.

🔧 Engineering Swarm

Skills that enforce code quality, testing, and review processes.

SkillDescriptionFull Docs
cm-tddUse before writing implementation code. Test-driven development protocol.View →
cm-debuggingSystematic root cause investigation for bugs, test failures, or unexpected behavior.View →
cm-quality-gateEnforces test gates, evidence-based verification, and frontend safety checks before deploy. No claims without evidence.View →
cm-test-gateComplete guide to setting up reliable test gates — stack detection, 4 core test files, script wiring, CI patterns.View →
cm-code-reviewFull review lifecycle — request reviews, handle feedback, complete branch integration.View →

⚙️ Operations Swarm

Skills that ensure safe deployment, secret management, and workspace isolation.

SkillDescriptionFull Docs
cm-safe-deployMulti-gate deploy pipeline with test gates, build verification, and rollback strategy.View →
cm-identity-guardVerifies project identity before git push, Cloudflare deploy, or Supabase operations. Prevents wrong-account deploys.View →
cm-git-worktreesCreates isolated git worktrees for feature work. Smart directory selection and safety checks.View →
cm-terminalEnforces progress logging, output reading, and error-stop for every terminal command.View →

🎨 Product Swarm

Skills for planning, design, documentation, and bootstrapping new projects.

SkillDescriptionFull Docs
cm-planningRequired before any creative or multi-step task. Explores intent, requirements, and design before implementation.View →
cm-brainstorm-ideaMulti-dimensional evaluation using Design Thinking + TRIZ + Double Diamond. For existing product enhancements.View →
cm-ux-masterUltimate UI/UX design intelligence — 48 UX Laws, 37 Design Tests, Figma & Stitch integration.View →
cm-ui-previewAI-powered UI generation using Google Stitch MCP. Stitch Build Loop + Prompt Optimization.View →
cm-dockitKnowledge systematization engine — generates tech docs, SOP guides, API references as VitePress or Markdown.View →
cm-readitTurn any website into an audio-enabled experience — TTS, MP3 player, Voice CRO. Zero dependencies.View →
cm-project-bootstrapNew project scaffolding — design system, staging+production, i18n, SEO, AGENTS.md, test infrastructure, 8-gate pipeline.View →

📈 Growth Swarm

Skills for content creation, marketing automation, and conversion tracking.

SkillDescriptionFull Docs
cm-content-factorySelf-learning content engine with StoryBrand, Cialdini, SUCCESs, Hook Model, JTBD, CRO frameworks.View →
cm-ads-trackerComplete conversion tracking — Facebook/Meta Pixel + CAPI, TikTok Pixel, Google Ads, GTM container.View →

🎯 Orchestration Swarm

Skills that coordinate multi-skill workflows, working memory, and autonomous execution.

SkillDescriptionFull Docs
cm-executionAutonomous execution with 3 modes: batch, subagent-per-task, or parallel dispatch.View →
cm-continuityWorking memory protocol — 4-Tier Memory System with context persistence, Ebbinghaus decay, and scope-filtered learnings across sessions.View →
cm-deep-searchSemantic memory power-up — detects large codebases and suggests qmd for local BM25 + Vector + LLM re-ranking search.View →
cm-skill-chainCompose skills into automated pipelines with progress tracking and auto-detection.View →
cm-skill-masteryMeta-skill — when to invoke skills, how to create new skills, and skill discovery.View →
cm-safe-i18nMass i18n conversion with multi-pass batching, parallel dispatch, 8 audit gates. Battle-tested.View →
cm-how-it-workComplete guide to Cody Master — from idea to deploy. Read this first if you're new.View →
cm-exampleTemplate demonstrating the universal skill format compatible with all platforms.View →

Creating Your Own Skills

Every skill follows the universal format:

yaml
---
name: my-custom-skill
description: "One line — when to use this skill"
---

# Skill Name

> One-line summary

## When to Use
[Trigger conditions]

## Procedure
[Step-by-step instructions]

## Rules
[Do's and Don'ts]

See cm-example for a complete template, and cm-skill-mastery for the meta-guide on skill creation and discovery.

Transparency

Every skill in this library is fully open. Click "View →" on any skill above to read its complete SKILL.md content — nothing is simplified, summarized, or hidden. We believe the best way to trust AI automation is to see exactly what instructions the AI follows.

Open Source AI Agent Skills Framework