graph-mcp-java-gen
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| GRAPH_BACKEND | No | Backend for the graph catalog. Defaults to the local JSON fixture. Set to 'neo4j' to use a live Neo4j database. | fixture |
| OPENAI_API_KEY | No | API key for OpenAI, required to enable the multi-agent NLP tool (generate_java_test_nlp). |
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 | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| get_fixture_metadataA | Return provenance and version information for the active graph fixture. |
| search_graphB | Search the bounded synthetic graph by class, qualified name, or method. |
| generate_java_testD | Generate Java only when graph grounding, syntax, contract, and safety pass. |
| generate_java_test_from_intentC | Parse a bounded natural-language request, generate Java, and enforce all gates. |
| generate_java_test_nlpB | Multi-agent pipeline: LLM parses free-form NL, generates Java, LLM reviews it. Agent 1 (LLMIntentParser) normalises natural language to a structured intent. The deterministic generator produces graph-cited, validated Java source. Agent 2 (ReviewAgent) audits the result against a six-item checklist. Requires OPENAI_API_KEY in the environment. |
| validate_java_sourceC | Validate bounded Java source against syntax, graph, contract, and safety gates. |
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 6 tools
The tools are mostly distinct: metadata retrieval, graph search, three variations of Java test generation, and validation. The three generate_java_test* tools could be confused, but their descriptions clearly differentiate by input type (structured, bounded NL, free-form NL with review). Minor overlap exists but descriptions resolve ambiguity.
All tool names follow a consistent verb_noun pattern in snake_case: get_fixture_metadata, search_graph, generate_java_test, generate_java_test_from_intent, generate_java_test_nlp, validate_java_source. The naming is predictable and uniform throughout the set.
With 6 tools, the server is well-scoped for its purpose of generating and validating Java tests from a synthetic graph. Each tool serves a clear role in the workflow, and the count feels appropriate—not too few to be inadequate, not too many to be unwieldy.
The tool surface covers the full lifecycle: metadata retrieval, graph search for grounding, three generation pathways (including NL-based variants), and validation. There are no obvious missing operations for the stated purpose; the server provides a complete pipeline from context gathering to output validation.