Loading ToolBench
A MCP server for asking questions and getting context on John Larkin (johnlarkin1 GH profile, johnlarkin1.github.io)
All 11 tools lack input parameter descriptions entirely. While tool names follow verb_noun convention (get_*, search_*) and descriptions exist (avg 65 chars), parameter documentation is completely absent. The search_info tool has a 'query' parameter with type and description, but the other 10 tools have empty input schemas ({}). No output schemas are documented. Error handling and recovery guidance are absent. This server reads like a personal resume API with minimal LLM-optimization for agent reasoning.
Return identifiers for all available content resources.
Return the extended biography content.
Return contact instructions for reaching John.
Return metadata describing the MCP server and resume freshness.
Return the curated list of noteworthy projects.
Return the full resume content as Markdown.
Return the current skills overview.
10 of 11 tools have empty input schemas ({}). No parameter descriptions, types, or constraints. LLMs cannot reason about what inputs these tools accept.
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields from responses.
Tool descriptions are generic and brief (avg 65 chars, baseline 194 chars). Missing WHEN to use, prerequisites, and what data is returned. E.g., 'Return the full resume content as Markdown' lacks context about format, length, or use case.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 43 | 2026-07-28+ | v2 |
Return collegiate tennis career information including awards and match records.
Return detailed work experience and employment history.
Return server health status and resource availability.
Return a formatted summary of resources matching the query string.
No error handling or recovery guidance. Tools do not document failure modes, retryability, or what the LLM should do if a call fails.
search_info is the only tool with a parameter schema, but it lacks output schema documentation. Agents cannot know what fields the search result contains.