m1nd

m1nd is an open-source, Rust-based code intelligence harness and Model Context Protocol (MCP) server created by maxkle1nz. Designed to provide structural and temporal context to coding agents, m1nd constructs detailed code graph representations with temporal, topological, and dynamic runtime metadata. Operating primarily as an MCP tool provider, m1nd exposes dozens of analysis tools (including over 68 discrete tool handlers) that AI agents can invoke over MCP interfaces. It parses Git history for co-change relationships, tracks taint propagation across architectural boundaries, matches structural clones, and ingests OpenTelemetry traces to overlay production execution data onto static code graphs. m1nd differentiates itself through its deep graph-based analysis engines and runtime tracing overlays, moving beyond basic AST parsing or vector search. By treating repository history as a 4D graph and combining it with taint analysis and community detection algorithms, it equips external LLM agents with rigorous structural context for refactoring and security analysis. The project is open-source under development on GitHub, implemented in Rust with a strict test harness and gate architecture.

Tags: ai agent, cli, codebase, coding agent, memory, OpenSource

Visit m1nd

Key features

  • 4D Git Graph Analysis — Parses repository Git history to map co-change frequencies and inject temporal ghost edges into the structural code graph. - Taint Tracing & Boundary Fuzzing — Analyzes data flow paths and detects boundary crossing to identify security vulnerabilities and untrusted input paths. - Topological Structural Twins — Calculates cosine similarity across code topology to discover isomorphic structures and duplicate logic. - OpenTelemetry Runtime Overlay — Ingests OTel traces to overlay dynamic execution heat and call frequency onto static repository graphs. - Intent-Driven Refactoring — Uses graph community detection and counterfactual extraction to formulate multi-file architectural refactor plans. - Extensive MCP Tooling — Exposes an expansive suite of 68+ protocol tools for MCP-compliant AI agents and terminal harnesses.

Use cases

Use Case: Architectural Refactoring with Temporal Graph Data A developer uses an MCP-connected coding agent to plan cross-module refactoring. The agent invokes m1nd's ghost_edges and refactor_plan tools to inspect historical co-change patterns and community boundaries, ensuring related modules are modified consistently without breaking architectural seams.

Use Case: Security Boundary Taint Analysis A team audits untrusted inputs across a distributed codebase. The agent calls m1nd's taint_trace tool to propagate data flow paths across structural boundaries and identify unprotected endpoints.

Use Case: Production Hotpath Code Optimization Developers overlay real-world usage onto static code by ingesting OpenTelemetry traces via the runtime_overlay tool, directing the AI agent to optimize bottlenecks in high-throughput paths.

Who it is for

m1nd is designed for systems engineers, AI agent tool builders, and senior developers looking to provide rich structural, temporal, and runtime context to LLM agents via MCP. It is less suitable for developers seeking a self-contained chat GUI or simple regex/AST search utilities without an MCP-capable host.

Install guide

Installation m1nd is built in Rust. Clone the repository and compile the binaries using Cargo: bash git clone https://github.com/maxkle1nz/m1nd.git cd m1nd cargo build --release

First Run Run the test suite to verify the installation: bash cargo test To run the project in MCP server mode, configure your MCP client (such as Claude Desktop or an MCP-compatible terminal agent harness) to execute the built binary: json { "mcpServers": { "m1nd": { "command": "/path/to/m1nd/target/release/m1nd", "args": ["serve"] } } }

Example Session Once configured in an MCP host, prompt your agent to execute code graph queries: text Use m1nd to trace temporal ghost edges and identify files frequently modified together with src/server.rs over the last 6 months.

Part of CLI Coding Agents

Categories: AI Coding Agents, Code Assistants, Developer Tools

Related CLI agents

  • Mneme — Mneme is an open-source orchestration harness and memory governance framework developed by MnemeHQ designed to prevent architectural drift across the…
  • Flameox — Flameox is an open-source coding agent harness and orchestration framework maintained by morluto. Built to support hybrid development workflows across…
  • Agent Memory System — Agent Memory System is an open-source agent orchestration harness developed by RavByte-AI. It provides structured memory retention and retrieval mechanisms…
  • Grite — Grite is an open-source terminal coding agent harness developed by Neul Labs. Implemented in Rust, it provides a lightweight command-line…
  • GoodMemory — GoodMemory is an open-source agent memory orchestration harness and context management system developed by hjqcan. Designed to give AI agents…
  • pi-mem — pi-mem is an open-source memory harness and persistent context manager developed by Jo Inc for the Pi CLI coding agent.…
  • AI Tools
  • Categories
  • Industries
  • CLI Coding Agents
  • MCP Servers
  • MCP Categories