minga-graph
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| PROJECT_PATH | Yes | Path to the project whose graph the MCP server should serve, passed as the positional argument: `claude mcp add minga-graph -- minga-graph mcp /path/to/project`. The server reads graph.json from minga-out/ in that project (read-only, no network access). |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| graph_queryA | Return the part of the code knowledge graph relevant to a plain-language question (nodes, edges, file:line evidence). Use before grepping the codebase. |
| graph_explainB | Explain one node (class, function, file, table, heading): location, doc, every connection with its evidence. |
| graph_pathB | Shortest path between two nodes, hop by hop, with evidence for every edge. |
| graph_impactB | What depends on a node (transitive callers, importers, subclasses). Use before changing or deleting code. |
| graph_statsB | Size of the graph and when it was built. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 5 tools
Each tool targets a distinct operation: explain one node, path between two, transitive impact, graph metadata, and natural-language query. graph_explain and graph_query could be mildly confused since both return node/edge info, but the 'one specific node' vs 'plain-language question' framing helps separate them.
All five tools follow the same graph_<verb/noun> pattern (graph_explain, graph_path, graph_impact, graph_stats, graph_query). Predictable and easy to scan.
Five focused tools for a read-only graph knowledge service; each earns its place without redundancy or thinness.
Covers node inspection, traversal, dependency analysis, metadata, and semantic query—full read-only lifecycle for a graph service. A direct node-list/lookup-by-name operation could be a minor convenience gap, but graph_query largely compensates.