rustkit-ai

rustkit-ai — AI-Native Developer Tools in Rust

rustkit-ai is an open source organization I founded to build tools that solve the real friction points in AI-assisted development: context loss between sessions, token costs that add up fast, and the absence of local semantic search. All tools are MIT licensed, published on crates.io, and work out of the box with the major AI coding agents.

The tooling was born out of practical needs encountered during internal R&D work at Ippon Technologies, where the same problems — context loss, token costs, codebase intelligence — appeared repeatedly on real production codebases.

Tools

aimemo v0.1.11 Persistent memory for AI coding agents. Writes structured context to CLAUDE.md, .cursor/rules, .windsurfrules, or .github/copilot-instructions.md depending on the editor. One command to set up, zero configuration required. Works with Claude Code, Cursor, Windsurf, and GitHub Copilot.

cargo install aimemo
aimemo setup --claude

trimcp v0.1.3 MCP proxy that reduces LLM token costs by 60–90% through compression and caching of MCP tool outputs before they reach the model. Auto-detects Claude Code and Cursor MCP configurations.

cargo install trimcp
trimcp setup

tersify v0.5.1 Compresses code and text for LLMs by stripping noise — comments, blank lines, redundant whitespace — reducing tokens by up to 50% without losing meaning. Useful as a preprocessing step before injecting code into any LLM context.

cargo install tersify
tersify src/main.rs | pbcopy

semtree v0.1.5 Semantic code intelligence for Rust — tree-sitter parsing, embeddings, and RAG for multi-language codebases. Parses and indexes Rust, Python, JavaScript, TypeScript, and Go from a single unified architecture.

cargo install semtree-cli

semstore v0.1.0 Local semantic search for Rust applications — store text, search by meaning, no cloud API required. On-device embeddings that run anywhere Rust runs.

References

• rustkit-ai — Official website
• rustkit-ai — GitHub organization