AtCode
AI Code Knowledge Mapping System-Build a query-able knowledge graph for code warehouses through Tree-sitter AST parsing and Memgraph graph database, supporting cross-file structure understanding, call chain analysis, and AI Agent integration.
##Functional characteristics - **AST Knowledge Map **: Automatically parse the code warehouse and build a knowledge map at the function/class/module level - ** Graph search **: Support searching for functions, classes and methods by keywords, glob, and regular - ** Call chain tracing **: Visualize call relationships, dependence paths, and class inheritance levels - ** Paper analysis **: Integrate Hugging Face Daily Papers to automatically discover and build related code maps - **MCP integration **: Expose mapping capabilities to AI coding tools such as Claude Code and Codex through the MCP protocol - ** Multi-language support **: Python (Tier 1), JavaScript/TypeScript/Java (Tier 2), C++/Rust/Go (Tier 2) - ** Real-time incremental synchronization **: File monitoring + update of definition level differences to keep maps and code synchronized ###Usage method 1. access application 2. Click **Add Repository** 3. Enter the Git URL (such as `https://github.com/pallets/flash.git`) 4. Wait for the map construction to be completed (the progress can be viewed on the interface)
