latent-k
Stop wasting tokens - give your AI a map of your code
latent-k – persistent dependency map for AI coding assistants
Summary: latent-k maintains a persistent dependency map of your codebase that is automatically injected as context to AI coding assistants, reducing token usage and improving efficiency. It tracks imports, exports, and callers, syncing changes and supporting multiple languages.
What it does
latent-k auto-syncs dependency maps when files change and provides an MCP server for on-demand queries, supporting JS/TS, Python, Go, Rust, Java, PHP, Ruby, C#, and C/C++.
Who it's for
It is designed for developers using AI coding assistants like Claude Code and Gemini CLI who want to reduce redundant context exploration.
Why it matters
It minimizes wasted tokens by giving AI assistants a structured map of the codebase, speeding up code understanding across sessions.