A curated collection of 18 specialized AI agent personas for Claude Code subagents. Each agent provides expert-level guidance in specific domains—from technical architecture to business strategy.
Copy agent files to your Claude Code agents directory:
# For user-level agents (available across all projects)
mkdir -p ~/.claude/agents
cp agents/*.md ~/.claude/agents/
# For project-level agents (available in current project only)
mkdir -p .claude/agents
cp agents/*.md .claude/agents/Once installed, Claude Code will automatically invoke agents when appropriate. You can also explicitly request an agent:
> Use the technical-lead-archie agent to review my API design
> Have the code-quality-inspector agent check my recent changes
> Ask the github-commit-specialist agent to help commit these changes
View and manage installed agents:
/agents
| Agent | File | Model | Description |
|---|---|---|---|
| A.R.C.H.I.E. | technical-lead-archie.md |
sonnet | Technical lead for Cloud (AWS/GCP/Azure), Python, TypeScript, and Godot Engine. Provides architectural decisions, code reviews, and mentoring. |
| A.P.I. | api-design-specialist.md |
haiku | REST and GraphQL API design specialist. Covers resource modeling, versioning strategies, and OpenAPI documentation. |
| T.E.S.T. | test-strategy-architect.md |
haiku | Testing strategy architect for pytest, Jest, Vitest, Playwright, and Godot GUT. Designs test pyramids and fixes flaky tests. |
| Q.A.I. | code-quality-inspector.md |
haiku | Code review specialist analyzing correctness, readability, performance, security, and testability. |
| D.E.B.U.G. | debugging-detective.md |
haiku | Systematic debugging and root cause analysis. Investigates production issues, performance problems, and cryptic errors. |
| S.E.C. | security-auditor.md |
haiku | Security review and vulnerability detection. Identifies OWASP Top 10 issues, injection flaws, and auth weaknesses. |
| D.E.V.O.P.S. | devops-sre-specialist.md |
haiku | DevOps and SRE specialist for CI/CD, Kubernetes, Terraform, and infrastructure reliability. |
| G.I.T. | github-commit-specialist.md |
haiku | Git workflow expert. Creates conventional commits, organizes changes into atomic commits, and maintains repository hygiene. |
| D.O.C. | technical-documentation-specialist.md |
haiku | Technical writing specialist for API docs, architecture overviews, and code documentation. |
| C.A.T.S. | technical-solution-advocate.md |
haiku | Vendor evaluation and build-vs-buy analysis. Assesses TCO, lock-in risks, and technical proposals. |
| Agent | File | Model | Description |
|---|---|---|---|
| G.E.N.T. | project-manager-gent.md |
haiku | Project manager for agile teams. Facilitates stand-ups, sprint planning, and blocker resolution. |
| C.E.O. | ceo-strategist.md |
haiku | Business strategy and executive guidance. Covers market analysis, competitive positioning, and growth strategy. |
| C.F.O. | cfo-financial-strategist.md |
haiku | Financial planning, budgeting, unit economics, and fundraising strategy. |
| S.A.L.E.S. | sales-strategist.md |
haiku | Sales process optimization, pipeline management, and go-to-market strategy. |
| G.R.O.W.T.H. | marketing-growth-specialist.md |
haiku | Marketing strategy, growth tactics, and brand positioning. |
| C.S.M. | customer-success-specialist.md |
haiku | Customer success, retention strategy, and churn prevention. |
| H.R. | hr-people-strategist.md |
haiku | Talent acquisition, organizational design, compensation, and culture development. |
| I.N.V. | inventory-supply-chain-specialist.md |
haiku | Supply chain optimization, inventory management, and logistics strategy. |
A.R.C.H.I.E. (Architectural Counsel & Hands-on Implementation Expert)
model: sonnet
tools: Read, Edit, Bash, Grep, Glob, WriteUse when you need:
- Cloud architecture design (AWS, GCP, Azure)
- Python performance optimization (FastAPI, Django)
- TypeScript/React architectural guidance
- Godot Engine game development expertise
- Code review with educational explanations
Example invocation:
> My Flask API is taking 3 seconds to respond. Use technical-lead-archie to optimize it.
A.P.I. (Application Programming Interface Specialist)
model: haiku
tools: Read, Grep, GlobUse when you need:
- RESTful API resource modeling
- GraphQL schema design
- API versioning strategies
- OpenAPI/Swagger documentation
- REST vs GraphQL trade-off analysis
Example invocation:
> Use api-design-specialist to design pagination for our user list endpoint.
T.E.S.T. (Test Strategy Architect)
model: haiku
tools: Read, Bash, Grep, GlobUse when you need:
- Test pyramid design for your project
- Flaky test diagnosis and remediation
- pytest fixture patterns
- Jest/Vitest/Playwright configuration
- Test coverage prioritization
Example invocation:
> Our E2E tests keep failing in CI. Use test-strategy-architect to fix them.
Q.A.I. (Quality Assurance Inspector)
model: haiku
tools: Read, Grep, GlobUse when you need:
- Systematic code review before PR submission
- Performance bottleneck identification
- Security vulnerability detection
- Best practices validation (PEP 8, ESLint)
- React/TypeScript anti-pattern detection
Example invocation:
> Use code-quality-inspector to review my new authentication module.
D.E.B.U.G. (Debugging Detective)
model: haiku
tools: Read, Bash, Grep, GlobUse when you need:
- Production bug investigation
- Stack trace analysis
- Performance degradation diagnosis
- Race condition identification
- Memory leak detection
Example invocation:
> We're getting random 500 errors in production. Use debugging-detective to investigate.
S.E.C. (Security Auditor)
model: haiku
tools: Read, Bash, Grep, GlobUse when you need:
- OWASP Top 10 vulnerability scanning
- Authentication/authorization review
- Injection flaw detection (SQL, XSS, CSRF)
- Secrets and credential scanning
- Dependency vulnerability assessment
Example invocation:
> Use security-auditor to review our payment processing code before launch.
D.E.V.O.P.S. (DevOps & SRE Specialist)
model: haiku
tools: Read, Bash, Grep, Glob, WriteUse when you need:
- CI/CD pipeline design (GitHub Actions, GitLab CI)
- Kubernetes deployment configuration
- Terraform infrastructure as code
- Docker containerization
- Monitoring and alerting setup
Example invocation:
> Use devops-sre-specialist to set up GitHub Actions for our monorepo.
G.I.T. (Git Specialist)
model: haiku
tools: Read, Bash, Grep, GlobUse when you need:
- Conventional commit message creation
- Atomic commit organization
- Git history cleanup before merge
- PR description writing
- Changelog generation
Example invocation:
> I've finished the OAuth feature. Use github-commit-specialist to commit my changes.
D.O.C. (Documentation & Organization Conduit)
model: haiku
tools: Read, Grep, Glob, WriteUse when you need:
- API reference documentation
- Architecture decision records (ADRs)
- README and onboarding guides
- Code docstrings and JSDoc
- System design documentation
Example invocation:
> Use technical-documentation-specialist to document our authentication flow.
C.A.T.S. (Customer Advocate for Technical Solutions)
model: haiku
tools: Read, Grep, Glob, BashUse when you need:
- Build vs buy analysis
- Vendor proposal evaluation
- Total Cost of Ownership (TCO) calculation
- Cloud provider comparison
- Lock-in risk assessment
Example invocation:
> Should we use Auth0 or build our own auth? Use technical-solution-advocate to analyze.
G.E.N.T. (Generative Engagement & Navigation Tool)
model: haiku
tools: Read, Grep, GlobUse when you need:
- Daily stand-up facilitation
- Sprint planning and retrospectives
- Blocker identification and escalation
- Project status reporting
- Meeting summaries and action items
Example invocation:
> Use project-manager-gent to prepare our sprint planning session.
C.E.O. (Chief Executive Strategist)
model: haiku
tools: Read, Grep, GlobUse when you need:
- Business strategy development
- Market opportunity analysis
- Competitive positioning
- Organizational prioritization
- Board presentation preparation
Example invocation:
> Use ceo-strategist to analyze our expansion into the European market.
C.F.O. (Chief Financial Strategist)
model: haiku
tools: Read, Grep, GlobUse when you need:
- Financial modeling and projections
- Unit economics analysis (CAC, LTV)
- Budgeting and resource allocation
- Fundraising strategy
- Cash flow management
Example invocation:
> Use cfo-financial-strategist to model our runway with different growth scenarios.
S.A.L.E.S. (Sales Strategist)
model: haiku
tools: Read, Grep, GlobUse when you need:
- Sales process optimization
- Pipeline management strategy
- MEDDIC/SPIN qualification frameworks
- Pricing strategy
- Go-to-market planning
Example invocation:
> Use sales-strategist to improve our enterprise sales qualification process.
G.R.O.W.T.H. (Marketing & Growth Specialist)
model: haiku
tools: Read, Grep, GlobUse when you need:
- Growth strategy and experimentation
- Content marketing planning
- SEO and acquisition channels
- Brand positioning
- Campaign performance analysis
Example invocation:
> Use marketing-growth-specialist to plan our product launch campaign.
C.S.M. (Customer Success Manager)
model: haiku
tools: Read, Grep, GlobUse when you need:
- Customer health scoring
- Churn prediction and prevention
- Onboarding journey design
- QBR preparation
- Expansion revenue strategy
Example invocation:
> Use customer-success-specialist to design our customer health score model.
H.R. (Human Resources Strategist)
model: haiku
tools: Read, Grep, GlobUse when you need:
- Recruiting process optimization
- Compensation framework design
- Organizational structure planning
- Performance review systems
- Culture development
Example invocation:
> Use hr-people-strategist to design our engineering leveling framework.
I.N.V. (Inventory & Network Virtuoso)
model: haiku
tools: Read, Grep, GlobUse when you need:
- Inventory optimization
- Demand forecasting
- Warehouse/3PL evaluation
- Supply chain cost analysis
- Logistics network design
Example invocation:
> Use inventory-supply-chain-specialist to optimize our reorder points.
All agents support a Task Tracking Workflow for complex, multi-step tasks. This enables:
- State persistence across memory compaction events
- Progress tracking for iterative work
- Parallel agent operation without file conflicts
- Recovery from session interruptions
For complex tasks, agents create three tracking files in the working directory:
| File | Purpose |
|---|---|
{task-id}-context.md |
Goal, constraints, and important context that must persist |
{task-id}-todos.md |
Checklist of items to process with progress markers |
{task-id}-insights.md |
Findings, recommendations, and results (iteratively updated) |
Use descriptive task identifiers (e.g., auth-audit, api-migration) to enable multiple agents to work simultaneously without conflicts.
Edit the YAML frontmatter to restrict or expand tool access:
---
name: code-reviewer
tools: Read, Grep, Glob # Read-only access
---Available tools: Read, Edit, Write, Bash, Grep, Glob, WebFetch, TodoRead, TodoWrite
Specify model preference in frontmatter:
---
model: sonnet # For complex reasoning tasks
model: haiku # For fast, simple tasks
model: opus # For most capable reasoning
model: inherit # Use main conversation's model
---- Create a new
.mdfile following the structure pattern - Define clear specialization (avoid overlap with existing agents)
- Include 3+ usage examples in the
descriptionfield - Test with
/agentscommand
This project is licensed under AGPL-3.0. Contributions and derivative works must be open-sourced under the same license.
- YAML frontmatter is valid (
name,description,model,color) - Description includes trigger phrase "Use this agent when..."
- Agent has clear, non-overlapping specialization
- Core Principles are domain-specific (not generic)
- Includes concrete examples from the domain
- Length is appropriate (~100-300 lines)
AGPL-3.0 - GNU Affero General Public License v3.0
Derivative works must be open-sourced under the same license.