About MCP
The Model Context Protocol is an open standard that enables AI assistants to securely access data and tools. MCP servers expose resources (data) and tools (functions) that AI models can use during conversations.Agents
MCP Developer
MCP server development, tool design, resource exposure, and transport implementation
- Mode:
subagent - Quality: 4.88/5 (Excellent)
- Tags: mcp, server, tools, resources, sdk, development
MCP Security Auditor
MCP server security reviews, OAuth, RBAC, and compliance frameworks
- Mode:
subagent - Quality: 4.12/5 (Good)
- Tags: mcp, security, oauth, rbac, compliance, audit
Usage Examples
Quality Stats
- Average score: 4.50/5
- 1 Excellent, 1 Good rating
- Total tokens: ~2,700 (avg ~1,335 per agent)
- Coverage: Development + security
Common Workflows
MCP Development Pack
MCP Development Pack
Both MCP agents for complete coverage:Includes: mcp-developer, mcp-security-auditor
Building an MCP Server
Building an MCP Server
- MCP Developer — Design server architecture and tools
- MCP Developer — Implement resources and transports
- MCP Security Auditor — Security review before deployment
- MCP Security Auditor — Compliance validation (if needed)
Securing an MCP Server
Securing an MCP Server
- MCP Security Auditor — Security architecture review
- MCP Security Auditor — OAuth/RBAC implementation audit
- MCP Developer — Implement security improvements
- MCP Security Auditor — Final validation
When to Use
Choose MCP Developer when...
Choose MCP Developer when...
- Building new MCP servers
- Designing tool schemas
- Implementing resource providers
- Integrating with existing APIs
- Working with MCP SDKs (TypeScript, Python)
- Setting up transport layers (stdio, SSE, HTTP)
Choose MCP Security Auditor when...
Choose MCP Security Auditor when...
- Auditing MCP server security
- Implementing OAuth flows
- Designing RBAC policies
- Ensuring compliance (SOC2, GDPR)
- Reviewing authentication/authorization
- Validating data access controls
MCP Concepts
Resources
Data that MCP servers expose to AI models:
- Files
- Database records
- API responses
- Real-time streams
Tools
Functions that AI models can invoke:
- Data queries
- API calls
- File operations
- Computation tasks
Security
Critical considerations for MCP servers:
- Authentication (OAuth, API keys)
- Authorization (RBAC, permissions)
- Data validation (input sanitization)
- Rate limiting
Transports
Communication protocols:
- stdio — Local processes
- SSE — Server-sent events
- HTTP — REST endpoints
Integration Examples
MCP Server + Backend
MCP Server + Backend
Combine MCP agents with backend specialists:
- API Architect — Design underlying API
- MCP Developer — Expose API via MCP
- Security Engineer — Security architecture
- MCP Security Auditor — MCP-specific security review
MCP Server + Database
MCP Server + Database
Database access via MCP:
- Database Architect — Design data model
- PostgreSQL Pro or SQL Pro — Optimize queries
- MCP Developer — Build query tools and resource providers
- MCP Security Auditor — Validate access controls
Resources
MCP Specification
Official protocol specification
MCP Documentation
Guides and examples
TypeScript SDK
Official TypeScript implementation
Python SDK
Official Python implementation