pforge
A declarative framework for building MCP servers using YAML configuration, generating optimized Rust code.
README
pforge
A declarative framework for building Model Context Protocol (MCP) servers using YAML configuration.
MCP Server
Registry Name: io.github.paiml/pforge
pforge is available in the Model Context Protocol (MCP) Registry. Install it via:
# Via Cargo (recommended)
cargo install pforge-cli
# Then run as MCP server
pforge serve
For Maintainers: See MCP Registry Publishing Guide for publishing instructions.
What is pforge?
pforge lets you define MCP servers in YAML instead of writing boilerplate code. It's built on top of pmcp (rust-mcp-sdk) and generates optimized Rust code from your configuration.
Quick example:
forge:
name: my-server
version: 0.1.0
transport: stdio
tools:
- type: native
name: greet
description: "Greet someone"
handler:
path: handlers::greet_handler
params:
name: { type: string, required: true }
Installation
# From crates.io
cargo install pforge-cli
# From source
git clone https://github.com/paiml/pforge
cd pforge
cargo install --path crates/pforge-cli
Quick Start
# Create new project
pforge new my-server
cd my-server
# Run the server
pforge serve
The scaffolded project includes a working example handler. Edit pforge.yaml to add more tools, then implement handlers in src/handlers/.
Handler Types
pforge supports four handler types:
- Native - Rust functions with full type safety
- CLI - Execute shell commands
- HTTP - Proxy HTTP endpoints
- Pipeline - Chain multiple tools together
See the book for detailed examples of each type.
Language Bridges
pforge provides language bridges for building MCP servers in your preferred language:
Deno/TypeScript Bridge
Build type-safe MCP servers using TypeScript and Deno with native performance:
import { PforgeBridge } from "https://raw.githubusercontent.com/paiml/pforge/main/bridges/deno/bridge.ts";
const bridge = new PforgeBridge();
bridge.register({
name: "greet",
description: "Greet a user by name",
handler: (input: { name: string }) => ({
success: true,
data: { message: `Hello, ${input.name}!` },
}),
});
const result = await bridge.execute("greet", { name: "Alice" });
console.log(result.data);
Features:
- Type-safe handler definitions with TypeScript generics
- Runtime schema validation with no external dependencies
- O(1) handler lookup performance
- Both sync and async handler support
- 74+ tests passing with full quality gates
Documentation: bridges/deno/README.md
Coming Soon
- Python bridge (FFI-based with asyncio support)
- Go bridge (CGo-based with goroutine support)
- Node.js bridge (Native addon with N-API)
Documentation
- Book - Complete guide with examples and comparisons
- Architecture - Technical design details
- User Guide - Usage guide
- Implementation Status - Current project status
- CLAUDE.md - Development workflow for contributors
Examples
- hello-world - Minimal native handler example
- calculator - Math operations with tests
- rest-api-proxy - HTTP handler examples
Project Status
Version: 0.1.2
Published crates:
pforge-config- Configuration parsingpforge-macro- Procedural macrospforge-runtime- Core runtime (depends on pmcp)pforge-codegen- Code generationpforge-cli- CLI tool
Test results: 120+ tests passing (90+ unit/integration, 12 property-based, 8 quality gates, 5+ doctests)
See IMPLEMENTATION_STATUS.md for detailed progress.
Development
# Run tests
cargo test --all
# Run quality gates
make quality-gate
# Watch mode
make watch
# Build release
make build-release
See CLAUDE.md for full development workflow.
Architecture
pforge is built as a framework on top of pmcp (rust-mcp-sdk):
┌─────────────────────────────────┐
│ pforge (Framework Layer) │
│ - YAML → Rust codegen │
│ - Handler registry │
│ - State management │
└─────────────────────────────────┘
↓
┌─────────────────────────────────┐
│ pmcp (Protocol SDK) │
│ - MCP protocol implementation │
│ - Transport handling │
└─────────────────────────────────┘
When to use pmcp directly: You need fine-grained control over MCP protocol details or want to avoid code generation.
When to use pforge: You want declarative configuration and rapid MCP server development with less code.
Contributing
Contributions are welcome. Please:
- Read CLAUDE.md for development standards
- Check ROADMAP.md for current priorities
- Ensure tests pass:
cargo test --all - Ensure quality gates pass:
make quality-gate
All commits are validated by pre-commit hooks that check code formatting, linting, tests, complexity, coverage, and markdown link validity (using pmat validate-docs) to prevent broken documentation links.
License
MIT - see LICENSE
Acknowledgments
Built on pmcp by Pragmatic AI Labs.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.