Skip to main content
Glama

Version Platform Engine License


🖼️ Preview: 3D Interactive Obsidian-Style Graph Visualizer


Related MCP server: Graft

🚀 Overview

CodeGraph v2.0 is a next-generation static codebase architecture engine powered by the ultra-fast TokenVector (.tkv) runtime. CodeGraph transforms your entire repository into a deterministic Knowledge Graph and exposes it directly to AI Coding Assistants via the Model Context Protocol (MCP) standard.

By providing 100% deterministic structural links (function calls, class inheritance, module imports), CodeGraph completely eliminates AI hallucinations, enabling confident and precise codebase refactoring.


⚡ Real-World Benchmarks & Competitive Edge

📊 Benchmark Comparison

Metric

CodeGraph v2.0 (TokenVector)

Vector Search / Legacy AST Tools

Key Advantage

Scan Speed (800+ Files)

< 0.8 Seconds

12 - 45 Seconds

15x - 50x Faster. Zero parsing overhead.

Link Accuracy

100% Deterministic

Probabilistic

0% Hallucination. 2-level reverse dependency tracing.

Memory Footprint (RAM)

~ 8 MB RAM

500 MB - 2 GB RAM

Ultra-lightweight. No embedded graph DB required (Neo4j, KùzuDB).

Token Optimization

Saves 92% Tokens

Full repo dumping

Surgical context packaging (get_optimal_context).

Security & Privacy

100% Local & Offline

Uploads code to Cloud Vector DB

Source code never leaves your local machine.


🌟 5 Core Competitive Advantages

  1. 🎯 Cascading Impact Analysis (analyze_impact): Before an AI modifies a function or class, CodeGraph traverses 2 levels of reverse dependencies (Level 1 & Level 2) to warn: "Modifying this function will break Class B in file X and Unit Test C in file Y".

  2. 📦 Optimal Context Packaging (get_optimal_context): Automatically computes structural dependencies around target edits and packages them precisely within the AI's token limit, eliminating irrelevant noise.

  3. 🕸️ Interactive 3D Obsidian-Style Graph: Automatically exports an interactive 3D graph visualizer built with D3.js, featuring community color clustering and node degree sizing.

  4. 🔑 Dual-Layer Licensing: Combines seamless 24/7 activation via the Lemon Squeezy API with a fallback set of 100 permanent Offline Master Keys (valid through 2099-12-31) for enterprise and offline air-gapped environments.

  5. 🎁 Zero-Friction 30-Day Free Trial: Instant 30-day trial automatically activated upon installation—no sign-up or credit card required.


📥 Licensing & Downloads

👉 Download Release

👉 Purchase Official License Key on Lemon Squeezy

(Each license key supports simultaneous activation on up to 2 devices).


📖 Quick Start & Activation Guide

  1. Extract the CodeGraph2.0_Release archive.

  2. Double-click CodeGraph_v2_Launcher.bat to analyze the repository and generate the 3D graph view.

  3. Register the MCP server in your AI Editor's mcp_config.json:

    {
      "mcpServers": {
        "CodeGraph": {
          "command": "C:\\CodeGraph2.0_Release\\2.code\\tools\\pytok.exe",
          "args": ["C:\\CodeGraph2.0_Release\\2.code\\CodeGraph_MCP.exe"]
        }
      }
    }
  4. For comprehensive usage instructions, refer to USER_GUIDE.md.


© 2026 CodeGraph. All rights reserved. TokenVector Engine Technology.

Related MCP Connectors

Related MCP Servers

  • A
    license
    A
    quality
    A
    maintenance
    An MCP server and CLI tool that transforms codebases into AI-ready context through semantic search, call graph analysis, and incremental indexing. It enables AI assistants to perform hybrid vector and keyword searches to understand complex repository structures and cross-file relationships.
    35
    200 npm
    1
    MIT
  • A
    license
    B
    quality
    D
    maintenance
    Local-first codebase context engine that parses code into a ranked dependency graph and serves it to AI tools via MCP for deep structural understanding.
    5
    15 npm
    1
    MIT
  • A
    license
    A
    quality
    A
    maintenance
    Enables AI coding assistants to understand codebase architecture in real time by parsing source code into a relationship graph and exposing call chains, dependencies, class hierarchies, and conventions via MCP tools.
    14
    4
    MIT
  • F
    license
    Not graded
    quality
    B
    maintenance
    Builds a semantic knowledge graph of C++ code and exposes 9 MCP tools for AI assistants to search classes, functions, inheritance, callers, callees, overrides, and more.
    3
    -