Generate MCP servers
in seconds
mcp-forge analyses any API, schema or codebase and writes a fully working MCP server ready to plug into Claude, Cursor, VS Code or any AI client.
What is the Model Context Protocol?
The Model Context Protocol (MCP) is an open standard created by Anthropic that allows AI assistants like Claude, Cursor or GitHub Copilot to call external tools and APIs directly during a conversation, without copy-pasting or manual integration.
An MCP server exposes a set of callable tools (functions, queries, mutations) that AI clients discover and invoke automatically. Writing one by hand requires understanding the protocol, structuring tool schemas and handling transport. mcp-forge does all of this for you in seconds.
Point mcp-forge at any OpenAPI spec, GraphQL schema, codebase,
CLI tool or website and get a ready-to-run server.py
that any MCP-compatible client can connect to immediately.
Instant generation
One command. A production-ready server in seconds.
Any source
OpenAPI, GraphQL, codebases, CLI tools, websites.
AI-enriched
Claude enriches tool names and descriptions automatically.
Ready to ship
Generated servers pass smoke tests and MCP handshake.
Get started in 2 minutes
# Install
pip install vulcai-mcp-forge-cli
# Generate from any API
mcp-forge run https://petstore3.swagger.io/api/v3/openapi.json
Supported sources
Automatic quality evaluation
Every generated server is scored on 6 criteria
Code structure
Readability, organisation, FastMCP idioms
Endpoint coverage
All tools present and correctly mapped
Auth & security
Env vars, headers, no hardcoded secrets
Description quality
Docstrings, tool names, argument docs
Spec fidelity
Types, body vs params, path parameters
Robustness
Timeout, error handling, default values
Servers scoring below 6/10 are automatically refined before delivery.
Generation returned 0 tools? Troubleshooting guide →