Available Agents (106 commands, 44 sub-agents, 53 skills)

Orchestrator (Single entry point)

Command	Mode	Usage
/assistant	Guide	Analyze → Recommend → Wait for confirmation
/assistant-auto	Automatic	Analyze → Execute the workflow directly

WORK-: Main Workflow (15)

Command	Usage
/work:work-explore	Explore and understand the code
/work:work-brainstorm	Structured ideation before specification (alternatives, questioning)
/work:work-specify	Create a functional specification (User Stories, criteria)
/work:work-clarify	Clarify spec ambiguities (targeted questions)
/work:work-plan	Plan an implementation (generates plan.md + tasks.md)
/work:work-commit	Create a clean commit
/work:work-pr	Create a Pull Request
/work:work-commit-push-pr	Full workflow: commit + push + PR
/work:work-team	Launch a coordinated team of agents (Agent Teams)
/work:work-quick	Quick workflow for trivial changes (skip full cycle)
/work:work-batch	Sequential execution of user stories from a PRD
/work:work-flow-feature	Full feature workflow
/work:work-flow-bugfix	Full bugfix workflow
/work:work-flow-release	Full release workflow
/work:work-flow-launch	Full product launch workflow

DEV-: Development (16)

Command	Usage
/dev:dev-tdd	TDD cycle + generate tests for existing code + test infra setup
/dev:dev-debug	Debug an issue (4-phase methodology)
/dev:dev-refactor	Guided refactoring + entropy reduction
/dev:dev-document	Document generation (PDF, DOCX, XLSX, PPTX)
/dev:dev-api	Create/document a REST, GraphQL, or tRPC API + versioning
/dev:dev-component	Create a complete UI component (or custom hook)
/dev:dev-error-handling	Error handling strategy
/dev:dev-react-perf	React/Next.js performance optimization
/dev:dev-mcp	Create MCP servers (Model Context Protocol)
/dev:dev-flutter	Flutter widgets and screens
/dev:dev-supabase	Supabase backend (Auth, DB, Storage, Postgres perf)
/dev:dev-neovim	Neovim/Lua plugins and config
/dev:dev-rag	RAG systems (Retrieval-Augmented Generation)
/dev:dev-design-system	Design tokens and component library
/dev:dev-prisma	Prisma ORM (schema, migrations, queries)
/dev:dev-ai-integration	LLM integration (OpenAI, Claude API)

QA-: Quality (13)

Command	Usage
/qa:qa-review	Thorough code review + naming analysis
/qa:qa-security	OWASP security audit
/qa:qa-perf	Performance analysis
/qa:wcag-audit	WCAG accessibility audit
/qa:qa-audit	Full quality audit
/qa:qa-chrome	Chrome visual tests (DOM debugging, responsive, captures)
/qa:qa-design	UI/UX audit (100+ web design rules, incl. responsive/mobile-first)
/qa:qa-automation	Test automation
/qa:qa-loop	Autonomous audit-fix loop with stop criteria
/qa:qa-mobile	Mobile app quality audit (Flutter)
/qa:qa-neovim	Neovim config audit (perf, keymaps)
/qa:qa-e2e	End-to-End tests (Playwright, Cypress)
/qa:qa-tech-debt	Identify/prioritize technical debt (incl. coverage + Kaizen)

OPS-: Operations (28)

Command	Usage
/ops:ops-hotfix	Urgent production fix
/ops:ops-release	Create a release
/ops:ops-gitflow	GitFlow <init|feature|release|hotfix> <action> (init, branch start/finish/list)
/ops:ops-deps	Audit and update dependencies
/ops:ops-docker	Dockerize a project
/ops:ops-k8s	Kubernetes deployment (manifests, Helm)
/ops:ops-vps	VPS deployment (OVH, Hetzner, DigitalOcean)
/ops:ops-migrate	Code/dependency migration
/ops:ops-ci	CI/CD configuration
/ops:ops-monitoring	Instrument code + deploy the observability stack (Prometheus/Grafana/Loki)
/ops:ops-grafana-dashboard	Create Grafana dashboards (templates, alerts)
/ops:ops-database	DB schema, migrations
/ops:ops-health	Quick health check
/ops:ops-env	Environment management
/ops:ops-load-testing	Load and stress tests
/ops:ops-cost	Costs: Claude Code tokens + cloud FinOps optimization
/ops:ops-backup	Backup/restore + disaster recovery (RPO/RTO, failover)
/ops:ops-infra-code	Infrastructure as Code (Terraform)
/ops:ops-secrets-management	Secure secrets management
/ops:ops-mobile-release	App Store / Google Play publishing
/ops:ops-serverless	Serverless deployment (Lambda, Vercel, CF Workers)
/ops:ops-vercel	Vercel configuration and deployment
/ops:ops-proxmox	Proxmox VE infrastructure (VMs, LXC, network, backup)
/ops:ops-opnsense	OPNsense configuration via Terraform (firewall, NAT, DHCP/DNS)
/ops:ops-deploy	Secure deployment with pre-deploy checklist
/ops:ops-rollback	Secure rollback procedure
/ops:ops-standup	Cross-repo morning briefing (commits, PRs, CI, blockers, priorities)
/ops:ops-ci-fix	Autonomous diagnosis and repair of failing CI/CD pipelines

DOC-: Documentation (5)

Command	Usage
/doc:doc-generate	Generate documentation
/doc:doc-changelog	Generate/maintain the changelog
/doc:doc-explain	Explain complex code
/doc:doc-onboard	Discover a codebase
/doc:doc-api-spec	Generate OpenAPI/Swagger spec

BIZ-: Business (9)

Command	Usage
/biz:biz-model	Business model, Lean Canvas
/biz:biz-mvp	Define the MVP
/biz:biz-pricing	Pricing strategy
/biz:biz-pitch	Create a pitch deck
/biz:biz-roadmap	Plan the roadmap + define OKRs
/biz:biz-launch	Full launch workflow
/biz:biz-competitor	Market study + competitive analysis
/biz:biz-personas	Create user personas
/biz:biz-research	User research

GROWTH-: Growth (9)

Command	Usage
/growth:growth-landing	Create/optimize a landing page
/growth:growth-seo	SEO audit
/growth:growth-analytics	Tracking and KPI setup
/growth:growth-app-store-analytics	App Store / Google Play metrics
/growth:growth-email	Email marketing templates
/growth:growth-ab-test	Plan A/B tests
/growth:growth-retention	Retention strategies
/growth:growth-localization	Multi-market localization strategy
/growth:growth-cro	CRO + funnel analysis + onboarding/activation

DATA-: Data (2)

Command	Usage
/data:data-pipeline	Design ETL/ELT pipelines
/data:data-modeling	Data warehouse modeling

LEGAL-: Legal (5)

Command	Usage
/legal:legal-docs	ToS, T&Cs, legal notices
/legal:legal-rgpd	RGPD/GDPR compliance
/legal:legal-payment	Payment integration
/legal:legal-terms-of-service	Terms of Service
/legal:legal-privacy-policy	Privacy Policy

Sub-Agents (44)

Claude automatically delegates to specialized agents (isolated context, restricted tools).

Concepts

Concept	Folder	Trigger	Context
Commands	.claude/commands/	Manual (/name)	Shared
Skills	.claude/skills/	Automatic	Shared
Agents	.claude/agents/	Auto-delegation	Isolated

Agents by domain

Domain	Agents	Models
Workflow & Doc	work-explore, work-quick, work-batch, doc-onboard, doc-generate, doc-changelog, doc-explain	haiku/sonnet
Quality & Audits	qa-audit, qa-loop, qa-security, qa-perf, wcag-audit, qa-claudemd, qa-e2e, qa-chrome	haiku/sonnet/opus (security, audit, loop)
Operations	ops-deps, ops-health, ops-docker, ops-deploy, ops-ci, ops-database, ops-monitoring, ops-infra-code, ops-proxmox, ops-opnsense, ops-cost	haiku/sonnet
Development	dev-debug, dev-flutter, dev-document, dev-tdd	haiku/sonnet (tdd, debug)
Business & Growth	biz-model, biz-competitor, biz-mvp, biz-personas, growth-seo, growth-analytics, growth-landing, growth-localization, growth-cro	haiku
Data	data-pipeline	haiku/sonnet
Legal	legal-rgpd, legal-payment, legal-privacy-policy, legal-terms-of-service	haiku/sonnet

Agent Configuration

Each agent defines: model (haiku/sonnet/opus), permissionMode (plan/default), disallowedTools, hooks, skills.

Model distribution: 10 haiku (trivial operations) / 29 sonnet (default) / 5 opus (critical reasoning: qa-security, qa-audit, qa-loop, dev-tdd, dev-debug). Read-only comprehension agents (work-explore, doc-onboard, doc-explain) run on sonnet — matching Anthropic's June 2026 change bumping the built-in Explore agent off haiku for exploration quality.