For cross-cutting documentation (debugging, database schema, environments, API versioning, build commands), see the parent README.
MCP (Model Context Protocol) server for ITM Platform. Exposes project management data and operations to AI assistants (Claude, ChatGPT, VS Code Copilot, Cursor, etc.) through a universal protocol.
See House Rules for coding conventions.
An MCP server acts as a bridge between AI assistants and ITM Platform's v2 REST API. Any AI client that supports the MCP standard can connect to this server and:
- Query projects, tasks, users, risks, issues, and budgets
- Run analytics via DataMart's GraphQL endpoint
- Execute workflow prompts (project status reports, risk analysis, workload reviews)
- (Phase 2) Create and update tasks, risks, and issues
The server is a thin proxy -- it translates MCP tool calls into v2 API calls and formats responses for AI consumption. All business logic stays in the microservices.
AI Client (Claude / ChatGPT / VS Code / Cursor)
│
│ MCP Protocol (JSON-RPC 2.0)
│ Transport: Streamable HTTP (prod) or stdio (local dev)
│
▼
┌──────────────┐
│ ITM.MCP │ TypeScript + @modelcontextprotocol/sdk
│ Server │ Tools, Resources, Prompts
└──────┬───────┘
│
│ HTTP (Token auth)
│
▼
┌──────────────┐ ┌──────────────┐ ┌──────────────────┐
│ ITM.API │────▶│ ITM.Tasks │ │ ITM.DataMart │
│ Gateway │────▶│ ITM.Account │ │ (GraphQL) │
└──────────────┘ └──────────────┘ └──────────────────┘
| Layer | Choice |
|---|---|
| Language | TypeScript |
| MCP SDK | @modelcontextprotocol/sdk |
| Transport | Streamable HTTP (production) + stdio (local dev) |
| Auth | Token pass-through to ITM v2 API |
| HTTP client | fetch or axios |
Full discovery, architecture decisions, phased capabilities, and implementation details:
Pre-development -- specification and architecture phase. No code yet.