explore-codebase

Featured

Navigate and understand codebase structure using the knowledge graph

AI & Automation 31,800 stars 2907 forks Updated 1 weeks ago MIT

Install

View on GitHub

Quality Score: 89/100

Stars 20%
100
Recency 20%
90
Frontmatter 20%
70
Documentation 15%
61
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

## Explore Codebase Use the code-review-graph MCP tools to find your way around the codebase. ### Steps 1. Call `get_architecture_overview_tool` for the community structure. Call `list_communities_tool`, then `get_community_tool`, only for the modules you need. 2. Call `semantic_search_nodes_tool` to find a function or class by name or keyword. 3. 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. 4. Call `list_flows_tool`, then `get_flow_tool` for one flow, to follow an execution path. 5. Call `find_large_functions_tool` to find oversized functions. 6. Call `list_graph_stats_tool` only when you need node, edge and language counts. ## Token Efficiency Rules - 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.

Details

Author
tirth8205
Repository
tirth8205/code-review-graph
Created
7 months ago
Last Updated
1 weeks ago
Language
Python
License
MIT

Integrates with

Similar Skills

Semantically similar based on skill content — not just same category