Codeledger

CodeLedger is a Model Context Protocol (MCP) server designed for software teams that want to supercharge their AI coding agents (e.g., Claude, Cursor, GitHub Copilot, Gemini CLI) with deterministic context selection and local, compounding memory. It analyzes code repositories, scores and selects only task-relevant files, records interaction outcomes, and supports an evolving institutional knowledge base. Integrating directly with MCP-compatible agents, CodeLedger offers tools for querying patterns, retrieving precise context bundles, and submitting coding outcomes—improving accuracy and efficiency while reducing wasted context window usage. It is aimed at engineering teams, devops, and AI agent users seeking persistent context, actionable memory, and measurable development quality.