knowledge-graph

You can find symbol references and analyze call graphs across large codebases by using the knowledge graph to connect files, classes, and functions. It exposes actions like find_symbol_references to locate where symbols are defined and used.

Semantic code discovery searches for code entities by intent rather than exact names. The knowledge graph maps relationships between files, classes, and functions, allowing you to search by meaning to locate relevant code structures.

You can analyze code structure and dependencies across projects by leveraging the knowledge graph. It exposes the get_code_structure action to explore file structures, modules, and dependency relationships within your repositories.

Yes, the knowledge graph supports MCP-enabled integration for scalable analysis across projects. You can use standardized parameters with actions like search_codebase to query code entities through the MCP server.

Use the search_codebase action with a query parameter to find functions like 'processPayment'. You can filter by node_types such as function to locate specific code entities and their definitions across the codebase.

You can fetch dependency graph statistics for modules using the get_graph_stats action. The knowledge graph connects files, classes, and functions, providing structural relationships and dependency data for your codebase analysis.