docfy-mcp
Provides tools for querying NestJS API endpoints documented with docfy, enabling AI agents to list and retrieve details of API endpoints without a browser.
Allows querying any OpenAPI 3.0/3.1 specification to list endpoints and get detailed endpoint documentation for use in code generation.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@docfy-mcplist endpoints for user management"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
docfy-mcp
An MCP (stdio) server that exposes an OpenAPI catalog as tools for coding agents (Claude Code, Cursor, etc.) to query while implementing a client — without leaving the editor to open a browser.
Works best with a catalog already documented via nestjs-docfy,
but accepts any valid OpenAPI 3.0/3.1 spec.
Tools
list_endpoints— lists all endpoints (method + path + summary). Accepts an optionalfilter(case-insensitive substring match on path/summary/tags).get_endpoint— takesmethod+path, returns the full "Copy for AI" text block (Purpose, Request, Parameters, Validation, Success Response, Error Responses).lint_spec— checks the loaded catalog for spec-quality issues: missing summary/description, missing tags, missing 4xx/5xx responses, undocumented response descriptions, duplicate operation IDs.diff_specs— compares the loaded catalog against another spec (pathorurl, e.g. a previous version from production or a git tag) and reports added/removed endpoints and breaking vs. informational field changes.contract_test— fires a real request at every endpoint (or a filtered subset) of an already-running server built from the loaded spec, and validates each live response against its declared schema. TakesbaseUrl, optionalheaders(repeatable"Name: value"strings, e.g. auth), and an optionalfilter.
Related MCP server: cod-api MCP Server
Usage
Published on npm — no need to clone or build:
# from a static file
npx docfy-mcp --spec ./openapi.json
# from a NestJS server running locally
npx docfy-mcp --url http://localhost:3000/docs-jsonThe JSON path isn't a fixed convention — it depends on what the project
passed to SwaggerModule.setup() (/api-json, /docs-json,
/swagger-json, ...). If --url returns 404, docfy-mcp probes the most
common paths on the same origin and suggests any that looks like a real
OpenAPI document.
For specs behind auth, repeat --header as many times as needed:
npx docfy-mcp --url https://api.example.com/api-json --header "Authorization: Bearer xyz"Why
--urldoesn't use swagger-parser's HTTP resolver:--urlfetches the spec directly instead of delegating to swagger-parser's resolver. By default, swagger-parser'ssafeUrlResolverblocks local/private URLs as an SSRF protection — which would break the most common use case here: pointing at a local NestJS dev server.
Local development
npm install && npm run build
node dist/cli.js --spec/--url ...
# or, via tsx:
npm run devRegistering as a local MCP server (Claude Code / Cursor)
Add a .mcp.json at the root of the project where the MCP client will run:
{
"mcpServers": {
"docfy": {
"command": "npx",
"args": ["-y", "docfy-mcp", "--url", "http://localhost:3000/docs-json"]
}
}
}Restart the MCP client — the list_endpoints and get_endpoint tools
should appear in the available tools list.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Connectors
Gateway between LLM agents and world data through eight tools and a bundled endpoint catalog.
Point Gecko at an OpenAPI spec; get first-call-correct, auth-hidden agent tools.
Turn any task into the right API calls: discover, evaluate, and integrate public APIs.
AI-callable tools for API mocking, testing, monitoring, security, and automation.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceExposes OpenAPI specifications as MCP tools, enabling AI assistants to explore and understand API structures, endpoints, schemas, and documentation through semantic queries.7MIT

cod-api MCP Serverofficial
FlicenseNot gradedqualityDmaintenanceExposes REST API endpoints defined in an OpenAPI Specification as MCP tools, allowing AI models to call them via the ModelContext Protocol.-- AlicenseNot gradedqualityDmaintenanceExposes OpenAPI endpoints as MCP tools, enabling LLMs to discover and interact with REST APIs through the MCP protocol.10MIT
- FlicenseNot gradedqualityDmaintenanceTurns any OpenAPI/Swagger spec into queryable tools for LLMs, enabling endpoint search, detail retrieval, and schema exploration.1-