Complete reference for all 203 local specialized AI agents organized by category with model assignments.
| Agent | Model | Description |
|---|
| backend-architect | opus | RESTful API design, microservice boundaries, database schemas |
| frontend-developer | sonnet | React components, responsive layouts, client-side state management |
| graphql-architect | opus | GraphQL schemas, resolvers, federation architecture |
| architect-reviewer | opus | Architectural consistency analysis and pattern validation |
| cloud-architect | opus | AWS/Azure/GCP infrastructure design and cost optimization |
| hybrid-cloud-architect | opus | Multi-cloud strategies across cloud and on-premises environments |
| kubernetes-architect | opus | Cloud-native infrastructure with Kubernetes and GitOps |
| service-mesh-expert | opus | Istio/Linkerd service mesh architecture, mTLS, and traffic management |
| event-sourcing-architect | opus | Event sourcing, CQRS patterns, event stores, and saga orchestration |
| monorepo-architect | opus | Monorepo tooling with Nx, Turborepo, Bazel, and workspace optimization |
| Agent | Model | Description |
|---|
| c-pro | sonnet | System programming with memory management and OS interfaces |
| cpp-pro | sonnet | Modern C++ with RAII, smart pointers, STL algorithms |
| rust-pro | sonnet | Memory-safe systems programming with ownership patterns |
| golang-pro | sonnet | Concurrent programming with goroutines and channels |
| Agent | Model | Description |
|---|
| javascript-pro | sonnet | Modern JavaScript with ES6+, async patterns, Node.js |
| typescript-pro | sonnet | Advanced TypeScript with type systems and generics |
| python-pro | sonnet | Python development with advanced features and optimization |
| temporal-python-pro | sonnet | Temporal workflow orchestration with Python SDK, durable workflows, saga patterns |
| ruby-pro | sonnet | Ruby with metaprogramming, Rails patterns, gem development |
| php-pro | sonnet | Modern PHP with frameworks and performance optimization |
| Agent | Model | Description |
|---|
| java-pro | sonnet | Modern Java with streams, concurrency, JVM optimization |
| scala-pro | sonnet | Enterprise Scala with functional programming and distributed systems |
| csharp-pro | sonnet | C# development with .NET frameworks and patterns |
| Agent | Model | Description |
|---|
| elixir-pro | sonnet | Elixir with OTP patterns and Phoenix frameworks |
| django-pro | sonnet | Django development with ORM and async views |
| fastapi-pro | sonnet | FastAPI with async patterns and Pydantic |
| haskell-pro | sonnet | Strongly typed functional programming with purity, advanced type systems, and concurrency |
| unity-developer | sonnet | Unity game development and optimization |
| minecraft-bukkit-pro | sonnet | Minecraft server plugin development |
| sql-pro | sonnet | Complex SQL queries and database optimization |
| Agent | Model | Description |
|---|
| database-optimizer | sonnet | Query optimization, index design, migration strategies |
| database-admin | sonnet | Database operations, backup, replication, monitoring |
| database-architect | opus | Database design from scratch, technology selection, schema modeling |
| Agent | Model | Description |
|---|
| conductor-validator | opus | Validates Conductor project artifacts for completeness, consistency, and correctness |
| Agent | Model | Description |
|---|
| test-automator | sonnet | Comprehensive test suite creation (unit, integration, e2e) |
| tdd-orchestrator | sonnet | Test-Driven Development methodology guidance |
| debugger | sonnet | Error resolution and test failure analysis |
| error-detective | sonnet | Log analysis and error pattern recognition |
| Agent | Model | Description |
|---|
| data-scientist | opus | Data analysis, SQL queries, BigQuery operations |
| data-engineer | sonnet | ETL pipelines, data warehouses, streaming architectures |
| Agent | Model | Description |
|---|
| ai-engineer | opus | LLM applications, RAG systems, prompt pipelines |
| ml-engineer | opus | ML pipelines, model serving, feature engineering |
| mlops-engineer | opus | ML infrastructure, experiment tracking, model registries |
| prompt-engineer | opus | LLM prompt optimization and engineering |
| vector-database-engineer | opus | Vector databases, embeddings, similarity search, and hybrid retrieval |
| llm-finetuning-architect | opus | Fine-tuning strategy, method/model selection, eval-gate ownership |
| llm-finetuning-training-engineer | sonnet | Dataset prep, Unsloth training runs, artifact export |
| llm-finetuning-eval-engineer | sonnet | Golden sets, judge calibration, checkpoint promotion verdicts |
| dgx-spark-ops-engineer | sonnet | DGX Spark (GB10/aarch64/CUDA 13) environment setup and diagnostics |
| Agent | Model | Description |
|---|
| docs-architect | opus | Comprehensive technical documentation generation |
| api-documenter | sonnet | OpenAPI/Swagger specifications and developer docs |
| reference-builder | haiku | Technical references and API documentation |
| tutorial-engineer | sonnet | Step-by-step tutorials and educational content |
| mermaid-expert | sonnet | Diagram creation (flowcharts, sequences, ERDs) |
| c4-code | haiku | C4 Code-level documentation with function signatures and dependencies |
| c4-component | sonnet | C4 Component-level architecture synthesis and documentation |
| c4-container | sonnet | C4 Container-level architecture with API documentation |
| c4-context | sonnet | C4 Context-level system documentation with personas and user journeys |
| Agent | Model | Description |
|---|
| business-analyst | sonnet | Metrics analysis, reporting, KPI tracking |
| quant-analyst | opus | Financial modeling, trading strategies, market analysis |
| risk-manager | sonnet | Portfolio risk monitoring and management |
| Agent | Model | Description |
|---|
| customer-support | sonnet | Support tickets, FAQ responses, customer communication |
| hr-pro | opus | HR operations, policies, employee relations |
| legal-advisor | opus | Privacy policies, terms of service, legal documentation |
Agents are assigned to specific Claude models based on task complexity and computational requirements.
| Model | Agent Count | Use Case |
|---|
| Fable | 0 | Longest-horizon autonomous work (tier above Opus; see criteria) |
| Opus | 55 | Critical architecture, security, code review, production coding |
| Sonnet | 71 | Complex tasks, support with intelligence |
| Haiku | 25 | Fast operational tasks |
| Inherit | 52 | Complex tasks where the user chooses the model at runtime |
Use when:
- Generating code from well-defined specifications
- Creating tests following established patterns
- Writing documentation with clear templates
- Executing infrastructure operations
- Performing database query optimization
- Handling customer support responses
- Processing SEO optimization tasks
- Managing deployment pipelines
Use when:
- Designing system architecture
- Making technology selection decisions
- Performing security audits
- Reviewing code for architectural patterns
- Creating complex AI/ML pipelines
- Providing language-specific expertise
- Orchestrating multi-agent workflows
- Handling business-critical legal/HR matters
Claude Fable 5 (model: fable) is the tier above Opus. It is opt-in in Claude Code
(v2.1.170+, never the default) and carries roughly 2.6× the effective cost of Opus
($10/$50 per MTok plus a ~30% heavier tokenizer), so reserve it for agents where Opus
demonstrably needs multiple attempts.
Use when:
- Running multi-hour autonomous sessions (large codebase migrations, overnight refactors)
- Executing end-to-end work from a single well-specified goal
- Coordinating long-lived parallel sub-agent workstreams
Avoid for:
- Security-analysis agents — Fable 5's cyber/bio safety classifiers fall back to
Opus 4.8 on that content anyway
- Anything Sonnet or Opus already completes in one pass
Other harnesses map fable to their top available model (see
authoring.md). Fable-tier agent bodies should state goals
and constraints rather than step-by-step scaffolding, and must not ask the model to echo
its internal reasoning.
The plugin ecosystem leverages Sonnet + Haiku orchestration for optimal performance and cost efficiency:
Sonnet: backend-architect (design API architecture)
↓
Haiku: Generate API endpoints following spec
↓
Haiku: test-automator (generate comprehensive tests)
↓
Sonnet: code-reviewer (architectural review)
Sonnet: incident-responder (diagnose issue, create strategy)
↓
Haiku: devops-troubleshooter (execute fixes)
↓
Haiku: deployment-engineer (deploy hotfix)
↓
Haiku: Implement monitoring alerts
Sonnet: database-architect (schema design, technology selection)
↓
Haiku: sql-pro (generate migration scripts)
↓
Haiku: database-admin (execute migrations)
↓
Haiku: database-optimizer (tune query performance)
Full-Stack Feature Development:
Sonnet: backend-architect + frontend-developer (design components)
↓
Haiku: Generate code following designs
↓
Haiku: test-automator (unit + integration tests)
↓
Sonnet: security-auditor (security review)
↓
Haiku: deployment-engineer (CI/CD setup)
↓
Haiku: Setup observability stack
Agents can be invoked through natural language when you need Claude to reason about which specialist to use:
"Use backend-architect to design the authentication API"
"Have security-auditor scan for OWASP vulnerabilities"
"Get performance-engineer to optimize this database query"
Many agents are accessible through plugin slash commands for direct invocation:
/backend-development:feature-development user authentication
/security-scanning:security-sast
/incident-response:smart-fix "memory leak in payment service"
To add a new agent:
- Create
plugins/{plugin-name}/agents/{agent-name}.md
- Add frontmatter with name, description, and model assignment
- Write comprehensive system prompt
- Update plugin definition in
.claude-plugin/marketplace.json
See Contributing Guide for details.