Loading connector details…
Loading connector details…
Choose a unique username to continue using AgentHotspot
by bitkaio • Uncategorized
Parses codebases into a persistent Neo4j graph and exposes MCP tools for AI agents to query structural code information.
Query cross-file function calls and dependencies efficiently.
Analyze codebase security aspects like auth guards and taint flows.
A persistent, structural representation of code for semantic queries.
Codesteward transforms any repository into a queryable structural graph using tree-sitter AST and stores it in Neo4j, enabling AI agents to efficiently answer complex codebase questions such as function call chains, auth guards, and dependencies. It supports multiple programming languages and offers four MCP tools including graph rebuilding, querying, augmentation, and optional taint analysis. This approach avoids repeated file scanning by providing a pre-built graph for cross-file and semantic queries, improving AI agent code understanding and interaction.