CookieGli

•
6
•

Streamlines code context for AI agents through high-density genome compression, rapid symbol indexing, and adaptive memory evolution.

CookieGli is a lightweight, zero-dependency Python toolkit designed to optimize codebase interaction for both human developers and AI agents. It tackles common challenges in large software projects such as context overload, slow code navigation, and inefficient impact analysis. By employing semantic code skeletonization, it significantly reduces code size while preserving core structure, making it ideal for LLM context windows. Furthermore, it provides lightning-fast symbol indexing, precise change impact analysis (Git Blast Radius), and an innovative Bayesian Darwinian memory pool that distills and stores practical lessons from error tracebacks for future use.

Características Principales

01Autonomous Error & Traceback Distiller for automatic lesson extraction
02Semantic Code Skeletonizer & Focus-Symbol Mode for 75-92% token reduction
03Project AST Genome to generate concise architectural maps (<600 tokens)
04B-Tree Inverted Symbol Index providing sub-0.1ms lookup latency
054 GitHub stars
06Git Blast Radius & Test Targeting to precisely analyze change impact

Casos de Uso

01Optimizing LLM context windows for code analysis and generation by compressing codebases.
02Facilitating rapid code navigation and symbol lookup in complex monorepos.
03Automating test suite targeting and change impact analysis for efficient CI/CD workflows.