Skip to main content

Open-source AI Code Intelligence for Every Codebase

Graph-powered analysis, instant insights, limitless extensibility.

CodePrism turns sprawling, multi-language repos into a navigable knowledge-graph your AI assistants can reason about. Rust-powered speed, MIT-licensed freedom.

Get StartedArchitecture Deep DiveAPI Reference❤️ Sponsor

CodePrism Server:

cargo install codeprism-mcp-servercodeprism-mcp-server --help

Mandrel Test Harness:

cargo install mandrel-mcp-thmoth --version

5 Languages & Counting

Rust, Python, JavaScript, TypeScript, Java – plus pluggable adapters.

100% AI-Generated Core

Built entirely with AI-first engineering practices.

< 50ms Query Latency

Lightweight graph store delivers sub-second answers.

MCP Native

First-class Model Context Protocol implementation for agent integration.

CodePrism Server

Graph-First Code Intelligence for AI Assistants

CodePrism transforms your codebase into an interactive knowledge graph that AI assistants can navigate and understand. Built with Rust for performance, it implements the Model Context Protocol (MCP) to provide standardized code intelligence across multiple programming languages. Available on crates.io and GitHub releases for instant installation.

  • 20+ Production Tools - From symbol search to complexity analysis
  • Multi-Language Support - JavaScript, TypeScript, Python, Rust, Java
  • Real-time Updates - Graph updates as your code changes
  • Enterprise Ready - Production-tested with 100% tool success rate
Get StartedLearn More

Ready in Minutes

# Install from crates.io
$ cargo install codeprism-mcp-server

# Start the MCP server
$ codeprism-mcp-server

# Example MCP tool schema
{
  "name": "analyze_complexity",
  "description": "Analyze code complexity",
  "inputSchema": {
    "type": "object", 
    "properties": {
      "target": {
        "type": "string",
        "description": "File or function to analyze"
      }
    }
  }
}

Get Started in 30 Seconds

# Install from crates.io
$ cargo install mandrel-mcp-th

# Create test specification
$ cat > filesystem-test.yaml << EOF
name: "Filesystem MCP Server"
server:
  command: "node"
  args: ["server.js"]
  transport: "stdio"
tools:
  - name: "read_file"
    description: "Read file contents"
EOF

# Run comprehensive tests
$ moth run filesystem-test.yaml

✅ Protocol Compliance: PASSED
✅ Tool Validation: PASSED  
✅ Error Handling: PASSED
📊 Coverage: 95% (19/20 tests)

Mandrel MCP Test Harness

Universal Testing Framework for MCP Servers

Mandrel (aka moth) is a comprehensive testing framework that validates any MCP server implementation for protocol compliance, functional correctness, and performance characteristics. Built on the official Rust MCP SDK for guaranteed accuracy. Install instantly from crates.io or GitHub releases.

  • Protocol Compliance - Full MCP 2025-06-18 specification validation
  • Performance Testing - Concurrent execution with configurable limits
  • Rich Reporting - HTML, JSON, and JUnit XML output formats
  • CI/CD Ready - GitHub Actions, GitLab CI, Jenkins integration
Quick StartDocumentation

Why Choose CodePrism?

Discover the powerful features that make CodePrism the ideal choice for code analysis

Advanced Parse Graph

Precise ASTs & cross-language symbol links stored in a lightning-fast graph.

AI-Optimised Metrics

Complexity, duplication & "hot-spot" scoring designed for LLM consumption.

Rust-Powered Speed

Native performance, tiny memory footprint, zero GC pauses.

APIGETPOSTPUTDEL{"}[...]{"}[...]

Open API Surface

REST, WebSocket, MCP JSON-RPC – integrate from any stack.

95%LOW

First-class Plug-ins

Add new languages or analytics in a single WASM bundle.

API

Built for OSS

MIT license, transparent roadmap, community-driven extensions.