Awesome GitHub RepositoriesCode Knowledge Graphs

Codegraph is a local codebase indexer and static analysis graph database that serves as a context provider for AI agents. It parses multiple programming languages into a searchable knowledge graph of symbols and dependencies, exposing these relationships to AI tools through the Model Context Protocol. The project distinguishes itself by aggregating relevant code snippets and symbol flows to reduce token usage for large language models. It automates the configuration of server settings and steering instructions across various AI agent platforms and command line editors to enable automatic code

Headroom is an AI gateway proxy and token optimizer designed to reduce the cost and latency of large language model interactions. It functions as an intermediary that intercepts traffic between clients and providers to apply context compression, request routing, and format translation. The system differentiates itself through a Model Context Protocol server implementation that delivers compression and retrieval tools to compatible AI hosts. It employs a content-aware compression pipeline and tiered importance scoring to trim redundant data from logs and tool outputs while preserving essential

CodeQL is a semantic code analysis engine and vulnerability scanning tool that treats source code as data. It utilizes a static analysis query language to define complex patterns and security vulnerabilities within a code graph database. The system represents source code as a relational database, enabling the execution of structural queries and data flow analysis. This approach allows for the detection of security flaws and coding errors across large-scale repositories. The tool provides capabilities for automated code auditing, static analysis security testing, and custom vulnerability dete

Qodo Cover est une plateforme de gouvernance d'ingénierie et un assistant piloté par l'IA conçu pour la revue de code automatisée et la génération de tests unitaires. Il utilise un graphe de connaissances de la base de code basé sur l'arbre de syntaxe abstraite (AST) pour mapper les dépendances et les relations architecturales, lui permettant d'analyser les pull requests et d'appliquer les standards de codage organisationnels. Le système se distingue par un pipeline d'analyse multi-agents qui effectue un raisonnement architectural et identifie des bugs au-delà du diff immédiat. Il dispose d'un serveur de protocole de contexte de modèle (MCP) pour exposer l'intelligence de la base de code à des outils externes et peut faire évoluer automatiquement les règles d'application en apprenant des décisions passées sur les pull requests. La plateforme fournit des capacités complètes pour la gestion des connaissances de la base de code, incluant l'exécution de recherches approfondies, le requêtage sémantique et le mapping des dépendances système. Elle inclut également des outils pour la génération itérative de tests unitaires afin d'augmenter la couverture de code et une remédiation automatisée pour appliquer des correctifs directement sur les pull requests. Les options de déploiement incluent le SaaS multi-tenant, le single-tenant ou des installations entièrement sur site (on-premises).

Potpie is an LLM codebase analysis platform and multi-agent orchestration framework designed to act as an AI software engineer. It parses repositories into a structured code knowledge graph, enabling AI agents to perform multi-hop reasoning, dependency tracing, and grounded technical analysis across large codebases. The system distinguishes itself through a spec-driven development framework where agents generate detailed technical specifications and architecture plans before implementing multi-file code changes. It utilizes a durable execution engine to coordinate specialized AI personas for

CodeGraphContext is a code graph indexer and visualization tool that analyzes source code to build graphs of functions, classes, and inheritance relationships. It functions as a Model Context Protocol server, providing a structured codebase index to AI assistants for context retrieval and natural language querying. The project features an interactive web interface that uses force-directed layouts to visualize code dependencies and symbols. To accelerate the setup of large projects, it supports the import of pre-calculated knowledge bundles for popular repositories. The system provides capabi