MCP server for Verse (UEFN) guidance, samples, and reference links
Server provides 9 read-only tools with consistent naming (verb_noun pattern: list-, get-, search-) and basic descriptions. All tools have input schemas with type definitions. However, descriptions are generic and lack LLM-optimization guidance (e.g., 'Return the available Verse guides' vs 'Retrieve a curated list of Verse programming guides; call this first to discover available topics'). Parameter descriptions are minimal or absent. Output schemas are not documented. No error handling guidance. Tools are well-composed (single responsibility) but lack the depth needed for production agent planning.
Return the full Markdown for a Verse guide
Return the full Verse source for a sample file
List known docs sources by id (used by search-docs)
Return the available Verse guides as resource links
Return the available Verse sample files
Search known docs sources by id with optional filtering
Search through Verse guides for a keyword
Descriptions lack LLM-optimization guidance. Generic descriptions like 'Return the available Verse guides' (34 chars) do not explain WHEN to call the tool vs alternatives, what structure is returned, or dependencies. Baseline for A+ tools: 50-200 chars with context.
Output schemas are not documented in tool definitions. LLMs cannot plan downstream calls or extract required fields (e.g., does list-guides return guide IDs, titles, descriptions?). Baseline: 100% of A+ tools document return types.
Parameter descriptions are missing or minimal. 'guideId' has a description, but 'query' in search-guides lacks context on format/length. 'maxMatches' lacks range constraints (e.g., 1-100). Baseline: 100% of A+ tool params have descriptions.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 56 | 2026-07-28+ | v2 |
Search Verse sample files for a query
Extract a topic section from the Verse syntax guide
No error handling guidance. Tools do not document what happens if a guideId is invalid, a search returns no results, or a file is missing. LLMs need recovery hints (e.g., 'If guide not found, call list-guides to discover valid IDs').
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). All tools are read-only and idempotent, but this is not declared in the schema. Agents cannot infer safety properties without explicit annotations.