The ultimate high-performance code context engine for AI-native development.
📖 Overview
NakshAstraMCP provides AI agents (Claude, Cursor, Antigravity, etc.) with deep, structural understanding of your local codebase. Using advanced AST parsing and high-speed semantic ranking, it delivers the exact context your developer tools need to solve complex problems across large projects.
🗺 Documentation Hub
Quickly navigate to detailed guides:
| 🚀 Setup Guide | 📖 User Guide | 🤖 Agent Guide |
|---|---|---|
| Install & Configure | Advanced Usage & Tips | Behavioral Guide for AI |
| 📜 LICENSE | 🛡 Security Policy | 💬 Discussions |
|---|---|---|
| *LICENSE * | Privacy & Data Safety | Community & Support |
| 🛠 Troubleshooting |
🏆 Assessment & Performance
- Overall Efficiency: 9.3 / 10 (Industry-leading structural context)
- Index/Search Latency: ~0.68ms p95 (Ultra-low latency on 10,000-file repos)
- Idle Memory: < 150 MB RAM
- Language Support: Deep AST integration for Python, JavaScript, TypeScript, Java, and Kotlin.
Lightning-fast hybrid search across multiple repositories.
✨ Key Features
- 🔍 Hybrid Multi-Repo Search — Indexed search across all your projects simultaneously.
- 🧠 Semantic Reranking — AI-powered results prioritized by conceptual relevance using FlashRank.
- 🌳 AST-Aware Analysis — Understands code structure (classes, functions, imports) natively.
- 🩺 Surgical Intelligence — High-precision tools (
read_file,find_symbol,find_references) for localized context retrieval. - 🛡️ Administrative Control — Full CLI control over server lifecycle (
stop,restart,logs). - 👁️ Real-Time Watcher — Changes are indexed instantly with mass-update protection.
- 🧩 Runtime Language Addons — Provision new Tree-sitter grammars at runtime.
- 📈 Visual Dashboard — Interactive Nebula Graph UI with health monitoring.
- 🧹 Operational Resilience — Built-in Memory Guard and WAL checkpointing for stability.
Detailed indexing statistics and server health monitoring.
🚀 Quick Start (Fast-Track)
1. Unified Installation
Requires uv. Install the secure binary wheel directly:
📥 Download v3.10.1 Secure Wheel (Windows)
uv tool install https://github.com/vijaytank/NakshAstraMCP-Docs/releases/download/3.0.0/nakshastramcp-3.10.1-cp313-cp313-win_amd64.whl --force
or if you get any errors try
python -m pip install .\nakshastramcp-3.10.1-cp313-cp313-win_amd64.whl
2. Register & Index
Initialize your workspace roots to build the local knowledge graph:
nakshastramcp start --workspace C:\path\to\your\project
3. Verification & Health
nakshastramcp status # Check indexing progress
nakshastramcp doctor # Perform full environment audit
💻 Hardware Tiers
NakshAstraMCP adapts to your system automatically:
| Tier | Specs | Capabilities |
|---|---|---|
| Minimal | 2 cores / 4 GB RAM | Core search engine |
| Recommended | 4 cores / 8 GB RAM | + Semantic reranking + High-performance indexing |
| Optimal | 8+ cores / 16 GB RAM | Full graph analysis + Deep reranking |
🌉 Multi-Client Connectivity
NakshAstraMCP supports concurrent sessions from multiple IDEs via the Dual Transport Bridge.
- Primary Host: Your main IDE (e.g., Cursor) starts the host session.
- HTTP Follower: Configure secondary tools (e.g., VS Code extension) to connect to the bridge:
- URL:
http://127.0.0.1:2102/mcp - Type:
streamable-http
- URL:
🛡 Security & Privacy
- 100% Local: No source code or indices ever leave your machine.
- Sensitive Data Detection: Integrated secret scanner prevents indexing of API keys.
- Sandboxed Execution: The engine only accesses registered workspace roots.
- Zero Telemetry: No usage data is collected. Fully offline operation.
© 2026 Vijay Tank. All rights reserved.