skills

Skills Catalog

A Copilot-first skill system with a portable core and a Copilot-exclusive orchestration layer.

Skills are organized in two tiers:

• Portable core — standard agentskills.io skills that work in Copilot CLI, Claude Code, Codex CLI, Hermes Agent, and any compatible runtime
• Copilot-exclusive layer — skills that use Copilot CLI's native GitHub MCP, Fleet mode, Background Delegation, and model routing — capabilities that depend on native Copilot features

Each skill is a subdirectory containing a SKILL.md file and can be installed directly into any compatible AI coding tool.

Cross-Platform Compatibility (agentskills.io)

All skills in this repository follow the agentskills.io open standard — the same spec used by Claude Code, Hermes Agent, Cursor, and Codex CLI.

What this means

• Portable: Skills written here can be used in Claude Code, Hermes, and other agentskills.io-compatible runtimes with minimal adaptation

• Standard format: The SKILL.md subdirectory layout (skills/<category>/<skill-name>/SKILL.md) is the canonical agentskills.io path

• Install from GitHub: Any agentskills.io-compatible tool can install skills directly:

  # Gemini CLI (native skills install command)
  gemini skills install github:drvoss/everything-copilot-cli/skills/development/fix-github-issue
  
  # Claude Code (manual copy — claude does not have a shell-level skills install command)
  # Copy the skill directory to your project's .claude/skills/ or user-level ~/.claude/skills/
  cp -r skills/workflow/commit-workflow ~/.claude/skills/workflow/

Cross-runtime compatibility table

Runtime	Compatible	Notes
GitHub Copilot CLI	✅	Native — all features available
Claude Code	✅	Skills portable; agent_type mapping differs
Codex CLI	✅	Skills work; no fleet/background equivalent
Hermes Agent	✅	Full agentskills.io spec support
Cursor	⚠️	Skill content reusable; different invocation model

Key differences when using in Claude Code

• Replace /fleet with Agent tool calls
• background-agent skill maps to & background subprocess
• sql tool state maps to ~/.claude/state.json or custom memory files
• Copilot-exclusive skills (skills/copilot-exclusive/) require adaptation

Installing Skills

GitHub Copilot CLI

# Install a single skill (project-level)
# Copy each skill directory to: .github/skills/<skill-name>/SKILL.md
# Example: skills/development/fix-build-errors/ -> .github/skills/fix-build-errors/

# Install user-level (available in all projects)
# Copy to: ~/.copilot/skills/<category>/<skill-name>/SKILL.md

Codex CLI

# Project-level
# Copy to: .agents/skills/<skill-name>/SKILL.md

# User-level
# Copy to: ~/.codex/skills/<skill-name>/SKILL.md

Gemini CLI

# Install directly from this repo using the skills command
gemini skills install github:drvoss/everything-copilot-cli/skills/<category>/<skill-name>

---

Skill Architecture (Three-Layer Model)

Skills in this repository fall into three functional layers, based on the harness research:

Layer	Role	Examples in this repo
Orchestrator	Coordinates teams, manages workflow, handles errors and synthesis	team-planner, fleet-parallel, orchestration/skills/*
Agent-Extending	Adds domain expertise to an agent — loaded when the agent needs specialized knowledge	Most development/, security/, testing/, documentation/ skills
External	Integrates external services and tools beyond the built-in CLI toolset	mcp-ecosystem, ai-visibility, github-pr-workflow

Why this matters: When composing workflows, load Orchestrator skills into the coordinating agent and Agent-Extending skills into specialist sub-agents. External skills can live in either layer depending on context.

---

Orchestration Hub

Some of this repository's strongest multi-AI capabilities live outside skills/ in the dedicated orchestration layer.

If you want to...	Start here
Delegate work to external AI tools	orchestration/skills/
Learn the core cross-AI patterns	orchestration/README.md
Fan out work across independent agents	fleet-parallel
Coordinate a specialist team from Copilot CLI	team-planner

---

Skill Categories

🔵 Copilot-Exclusive (copilot-exclusive/)

Skills that use features specific to GitHub Copilot CLI — background agents, fleet mode, plan mode, and GitHub integration.

Skill	Description
actions-debugging	Use when a GitHub Actions workflow fails — diagnose the run log, identify the root cause, and apply a targeted fix
agentic-engineering	Use when designing or decomposing a task for agent execution — applies 15-minute task units, eval-first loops, and explicit input/output contracts
autopilot-patterns	Use when you're ready to let Copilot execute a multi-step plan autonomously — configures guardrails and handles plan-to-autopilot transitions
background-agent	Use when a task is too long to block the current session — delegates to a cloud background agent via & or /delegate and opens a draft PR
context-prime	Invoke when starting a session (or resuming after a break) on a repo before making changes, to load live project context
cross-session-memory	Use when you need to recover prior context, decisions, or artifacts across sessions — search history and resume with the right state
fleet-parallel	Use when you need to run the same task across many files or contexts in parallel — triggers /fleet mode for batch operations
github-code-search	Use when you need real-world implementation examples or cross-repository context — search GitHub's global code index and reuse the results as grounded context
github-issue-triage	Use when you have a backlog of unorganized GitHub Issues — bulk-labels, prioritizes, and assigns at scale
github-pr-workflow	Use when creating, reviewing, or merging a PR — runs the full PR lifecycle through Copilot's built-in GitHub MCP
ide-switching	Use when moving between VS Code and Copilot CLI — transfers context so you don't lose state when switching environments
knowledge-curator	Use when recurring project lessons should be promoted from session context into durable instructions, docs, or decision records
mcp-builder	Use when you need to build a new MCP server — plan, implement, validate, reload, and test it end to end
mcp-ecosystem	Use when built-in tools don't cover a service you need — add a custom MCP server to extend Copilot CLI's capabilities
multi-model-strategy	Use when choosing which model to use for a task — routes to the best model based on task type and cost trade-offs
plan-mode-mastery	Use when you want structured, approval-gated planning before execution — switches to Plan Mode for complex multi-step tasks
scope-guard	Use when a task must stay inside a narrow writable boundary or when risky commands need an explicit stop rule
session-management	Use when a task spans multiple steps or sessions and needs structured state tracking via the built-in SQLite session database
stack-detector	Scan the project tech stack (package.json, config files, lockfiles) and recommend the most relevant skills and rules from this collection
task-intake-router	Use when a request arrives and the right execution path is unclear — routes work to the correct mode, agent type, model tier, and delegation pattern before implementation starts
ecosystem-intake	Use when evaluating an external repository, tool, or framework for adoption — applies the Adopt/Adapt/Reject filter and produces a structured intake report
team-planner	Use when a task is too large or multi-domain for a single agent — assemble a specialist team with SQL tracking and /fleet dispatch

🛠 Development (development/)

Core software development skills applicable to any project.

Skill	Description
api-and-interface-design	Define public APIs, CLIs, webhooks, or SDK contracts first so compatibility and validation stay intentional
code-review	Structured code review for quality and correctness
context-engineering	Optimize information delivery to AI agents — minimize noise, maximize signal
deprecation-and-migration	Safely remove old APIs and migrate to new patterns with a 3-phase process
fix-build-errors	Diagnose and fix build failures fast
fix-github-issue	Resolve a GitHub Issue end-to-end: read → locate → fix → test → PR
performance-optimization	Measure first, isolate bottlenecks, and prove that a performance change actually helps
pr-multi-perspective-review	Review PRs from 6 lenses: PM / Dev / QA / Security / DevOps / UX
refactor-clean	Remove dead code and simplify complex logic
skill-creator	Describe a workflow and get a properly structured SKILL.md scaffolded in minutes
source-driven-development	Verify framework and library APIs against current official docs before implementing
spec-driven-development	Write a technical spec before coding — defines interface, structure, and boundaries first
systematic-debugging	4-phase root cause analysis (reproduce → isolate → hypothesize → verify) before fixing
tdd-workflow	Test-driven development with red-green-refactor cycles

Combo Skills (activate when two technologies are used together):

Skill	Description
nextjs-prisma	Type-safe data fetching and Server Actions for Next.js App Router + Prisma projects
react-vitest	Component testing setup and patterns for React + Vitest projects
nestjs-prisma	PrismaService singleton, repository pattern, and unit testing for NestJS + Prisma

📝 Documentation (documentation/)

Skills for keeping docs accurate and up-to-date.

Skill	Description
add-to-changelog	Add versioned entries to CHANGELOG.md (Keep a Changelog format)
api-documentation	Generate and maintain API documentation
architecture-decisions	Document hard-to-reverse technical decisions as Architecture Decision Records (ADRs)
code-tour	Use when a user asks for a code tour, onboarding walkthrough, architecture tour, or PR tour — creates persona-targeted .tour files with real file and line anchors
doc-update	Sync documentation when implementation changes

🔒 Security (security/)

Skills for identifying and fixing security issues.

Skill	Description
agent-owasp-check	Audit an AI agent system against the OWASP Agentic Security Initiative Top 10
evaluate-repository	Score a repository 1-10 across 7 dimensions, including AI agent governance
input-validation	Validate and sanitize all user-supplied inputs
pr-security-review	Review a pull request for security vulnerabilities — auth, injection, secrets, and OWASP Top 10 before merging
secret-detection	Find and remove hardcoded secrets
security-bounty-hunter	Use when the goal is practical vulnerability discovery for responsible disclosure or bounty submission — focuses on remotely reachable, exploitable issues
security-scan	Run a security scan across the codebase

🧪 Testing (testing/)

Skills for improving test coverage and quality.

Skill	Description
browser-devtools	Verify frontend behavior at runtime — DOM validation, network inspection, performance profiling
e2e-testing	Write and run end-to-end tests
eval-harness	Use when you need to evaluate an LLM pipeline or AI feature systematically — sets up test cases, scoring rubrics, and pass/fail tracking
test-coverage	Analyze and improve test coverage
ux-audit	Use when a page or component feels confusing and you need a structured UX audit — applies 6 Krug-inspired usability checks and returns prioritized findings

🔄 Workflow (workflow/)

End-to-end development workflow skills — from planning to shipping.

Skill	Description
commit-workflow	Conventional commit messages with emoji, atomic splits, and pre-commit checks
cost-audit	Use when AI inference costs are growing unexpectedly or when comparing model choices by cost/quality ratio — produces an actionable cost reduction plan
council	Use when a decision has multiple credible paths and no obvious winner — convene a four-voice council (Architect, Skeptic, Pragmatist, Critic) for adversarial deliberation
deep-research	Use when a question requires comprehensive evidence gathering from multiple sources, systematic synthesis, and traceable citations
implementation-review	Use after a delegated implementation pass lands — compare the delivered diff against the original task spec and produce an actionable follow-up review
deployment-canary	Use when a release needs staged rollout checks, rollback thresholds, and explicit promote/hold decisions
release	Cut a versioned release: tag → GitHub Release → publish (npm/PyPI/Docker)
verification-before-completion	Prove a task is done with fresh command output before claiming success
sprint-workflow	Full sprint: Think → Plan → Build → Review → Test → Ship
security-audit	OWASP Top 10 + STRIDE threat modeling audit
sprint-retro	Data-driven retrospectives using /chronicle and git metrics
using-git-worktrees	Create isolated working directories for parallel branch work without cloning the repo repeatedly

📦 Product (product/)

Product management and strategy skills.

Skill	Description
create-prd	Generate a structured PRD grounded in Jobs-to-be-Done thinking
feature-prioritization	Impact × Confidence × Effort matrix with SQL tracking
launch-strategy	Product launch checklist: alpha → beta → GA
opportunity-solution-tree	Teresa Torres' OST framework: outcome → opportunity → solution → experiment
product-capability	Use when engineering needs testable, implementable technical requirements — capability spec with acceptance criteria, task breakdown, and traceability

📣 Content & Marketing (content/)

Extended skills (non-development tasks) — useful for marketers, content teams, and product growth roles in addition to developers. Classified separately from core development skills.

Skill	Description
ai-visibility	GEO optimization: llms.txt, AI crawler access, citation optimization
content-strategy	Keyword research, topic clusters, and content calendar planning
seo	Use when the user wants better search visibility, SEO remediation, schema markup, or Core Web Vitals improvements — distinct from ai-visibility (GEO)

---

Skill Format

All skills follow the agentskills.io spec:

skills/
└── <category>/
    └── <skill-name>/
        └── SKILL.md      ← required filename

SKILL.md frontmatter:

---
name: skill-name
description: What the skill does (shown in skill pickers)
metadata:
  category: <category>
---

Valid categories: development, testing, security, documentation, copilot-exclusive, workflow, product, content

If you add a new category, update the shared allowlist in scripts/skill-metadata.js and then update the catalog/docs in the same change.

---

Contributing

1. Create a directory: skills/<category>/<your-skill>/
2. Add SKILL.md with required frontmatter (name, description, metadata.category)
3. Follow the pattern of existing skills for content structure
4. Add to this README's catalog table
5. Run npm run validate && npm run lint:md && npm test before submitting a PR