Explore Codebase

by tirth82056b12d11625cbec3b6773e076cb3d136464fa90e5No licenseListed Oct 9, 2026Updated Oct 9, 2026

Navigate and understand codebase structure using the knowledge graph

AI-generated overview

Guides agents to navigate a codebase using code-review-graph MCP tools for structure and relationships.

What it does: This skill provides a step-by-step procedure for exploring an unfamiliar codebase through a knowledge-graph MCP toolset. It directs the agent to fetch architecture overviews, search for functions or classes, trace callers, callees and imports, follow execution flows, and find oversized functions. It also sets token-efficiency rules, such as calling a minimal-context tool first and preferring targeted queries over broad listings.
When to use it: Use it when an agent needs to understand the structure of a codebase before making changes or answering questions about it. It fits onboarding onto an unfamiliar repository or locating relevant modules, functions and execution paths.
Requirements: Requires access to the code-review-graph MCP tools named in the document (for example get_architecture_overview_tool, semantic_search_nodes_tool, query_graph_tool, list_flows_tool). No scripts are shipped; it is instructions only.

Use the code-review-graph MCP tools to find your way around the codebase.

Call get_architecture_overview_tool for the community structure. Call list_communities_tool, then get_community_tool, only for the modules you need.
Call semantic_search_nodes_tool to find a function or class by name or keyword.
Call query_graph_tool with callers_of, callees_of or imports_of to trace relationships. children_of on a file lists its functions and classes.
Call list_flows_tool, then get_flow_tool for one flow, to follow an execution path.
Call find_large_functions_tool to find oversized functions.
Call list_graph_stats_tool only when you need node, edge and language counts.

Call get_minimal_context_tool(task="<your task>") before any other graph tool.
Pass detail_level="minimal" wherever a tool accepts it. Use "standard" only when minimal is not enough.
Prefer a targeted query_graph_tool call over a broad listing call.
Budget: about five tool calls and 800 tokens of graph output per task.
Read the implementation and its tests before changing code. The graph narrows scope; it does not replace the source.

Source and attribution

License: No license

Content belongs to its original authors. SourceWeft indexes it from a public repository.