Claude Consult
Get expert help from Claude specialist agents during implementation. Each agent
has domain-specific skills pre-loaded and tools scoped to its role.
This is NOT a replacement for the review gates. The agent-loops skill
defines formal code review (specialist-review) and test audit
(diff-test-audit) gates that run after implementation. This skill is for
asking questions during implementation — before you've written code, while
you're writing it, or when you're stuck.
When to Use
- "How does this module work?" — before implementing
- "Does my approach match existing conventions?" — while implementing
- "Is this pattern secure?" — before committing to a design
- "Why is this test failing?" — when stuck
- "What do the testing standards say about X?" — before writing tests
- "Is this component accessible?" — during UI work
How to Invoke
claude -p --agent <agent-name> "<your question>"
All agents run headless (-p / print mode). They read the codebase, apply their
loaded skills, and return an answer to stdout. Capture or pipe as needed:
# Direct invocation
claude -p --agent debugger "Why is test_auth_flow failing in src/auth/tests.py?"
# Capture to variable
ANSWER=$(claude -p --agent security-auditor "Is the input validation in src/api/handler.py safe?")
# Save to file
claude -p --agent architect-reviewer "Does adding a cache layer between api and db match the existing architecture?" > /tmp/arch-advice.md
Agent Roster
Codebase Understanding
| Agent | Use when you need to... | Skills loaded |
|---|
search-specialist | Find where something is defined or called | codanna-codebase-intelligence |
architect-reviewer | Evaluate module boundaries and dependencies | system-design, api-design-patterns |
docs-architect | Understand existing documentation structure | documentation-production, code-explanation |
Code Quality & Patterns
| Agent | Use when you need to... | Skills loaded |
|---|
code-reviewer | Check if code follows project conventions | code-quality-workflow, testing-anti-patterns, secure-coding-practices |
refiner | Get suggestions for improving code quality | code-quality-workflow |
rest-expert | Design or check REST API patterns | api-design-patterns |
Security
| Agent | Use when you need to... | Skills loaded |
|---|
security-auditor | Full security review of a code path | owasp-top-10, secure-coding-practices, vibe-security, security-testing-patterns |
owasp-top10-expert | Check against OWASP Top 10 specifically | owasp-top-10, secure-coding-practices, vibe-security |
penetration-tester | Identify exploitable vulnerabilities | security-testing-patterns, owasp-top-10, threat-modeling-techniques |
jwt-expert | Review JWT/auth token handling | secure-coding-practices |
compliance-auditor | Check regulatory compliance | secure-coding-practices, owasp-top-10 |
Testing
| Agent | Use when you need to... | Skills loaded |
|---|
test-automator | Plan test strategy or check coverage approach | test-review, python-testing-patterns, testing-anti-patterns |
vitest-expert | Vitest-specific test patterns | testing-anti-patterns |
Debugging
| Agent | Use when you need to... | Skills loaded |
|---|
debugger | Investigate a test failure or error | systematic-debugging, root-cause-tracing |
Performance
| Agent | Use when you need to... | Skills loaded |
|---|
performance-engineer | Find bottlenecks in code paths | workflow-performance, python-performance-optimization, react-performance-optimization |
frontend-optimizer | Optimize frontend bundle/rendering | react-performance-optimization, design-system-architecture |
performance-monitor | Set up or check monitoring/observability | workflow-performance |
Frontend & UI
| Agent | Use when you need to... | Skills loaded |
|---|
ui-ux-designer | Review UI for usability and accessibility | ux-review, accessibility-audit, design-system-architecture |
component-architect | Design component APIs and composition | design-system-architecture, typescript-advanced-patterns |
interaction-designer | Review interaction patterns and states | ux-review, accessibility-audit |
state-architect | Choose or review state management approach | react-performance-optimization, typescript-advanced-patterns |
react-specialist | React-specific patterns and optimization | react-performance-optimization, typescript-advanced-patterns, testing-anti-patterns |
tailwind-expert | Tailwind styling and responsive design | design-system-architecture, accessibility-audit |
Language Specialists
| Agent | Use when you need to... | Skills loaded |
|---|
python-pro | Python idioms, async, decorators | async-python-patterns, python-testing-patterns, python-performance-optimization |
rust-pro | Ownership, lifetimes, traits | test-driven-development |
typescript-pro | Advanced types and type safety | typescript-advanced-patterns, testing-anti-patterns |
javascript-pro | JS/Node.js async and cross-runtime | testing-anti-patterns, typescript-advanced-patterns |
sql-pro | Query optimization and schema design | database-design-patterns |
Infrastructure
| Agent | Use when you need to... | Skills loaded |
|---|
cloud-architect | AWS/Azure/GCP architecture decisions | terraform-best-practices, kubernetes-deployment-patterns, gitops-workflows |
kubernetes-architect | K8s deployment and service mesh | kubernetes-deployment-patterns, kubernetes-security-policies, helm-chart-patterns, gitops-workflows |
terraform-specialist | IaC modules and state management | terraform-best-practices |
Data
| Agent | Use when you need to... | Skills loaded |
|---|
database-admin | DB operations, backup, replication | database-design-patterns |
database-optimizer | Slow queries, indexing, N+1 problems | database-design-patterns |
postgres-expert | Postgres-specific schemas and migrations | database-design-patterns |
Documentation
| Agent | Use when you need to... | Skills loaded |
|---|
mermaid-expert | Generate architecture/flow diagrams | documentation-production |
tutorial-engineer | Write tutorials and how-to guides | documentation-production, code-explanation |
technical-writer | Write clear technical documentation | documentation-production, code-explanation |
reference-builder | Generate API/config reference docs | documentation-production |
Coordination & Meta
| Agent | Use when you need to... | Skills loaded |
|---|
orchestrator | Break down complex tasks | task-orchestration, dispatching-parallel-agents |
prompt-engineer | Craft or improve prompts for AI features | writing-skills |
learning-guide | Understand a concept or pattern | code-explanation |
Operating Rules
1. Consult, don't delegate implementation
These agents advise — they don't write your code. Read the answer, apply it
yourself. The agent may suggest approaches, but you implement them.
2. Scope your questions
Bad: "Review everything in src/"
Good: "Is the input validation in src/api/handler.py:45-60 safe against injection?"
Narrow questions get better answers and use fewer tokens.
3. Don't replace review gates
Consulting security-auditor during implementation doesn't skip the formal
specialist-review gate afterward. The review gates exist for independent
verification — consulting during implementation is for getting guidance early.
4. Use the cheapest agent that works
For factual lookups (where is X, what does the standard say), prefer
search-specialist or the language-specific agents. Reserve
security-auditor, debugger, and performance-engineer for questions
that need deeper analysis.
5. One question per invocation
Each claude -p --agent call is a cold start. Don