MCP server for searching and retrieving verified public talent profiles from Prompthaus across 18 career pathways
Two tools with complete schemas and descriptions, but output documentation is missing and error handling lacks recovery guidance. talent_search has good parameter constraints (limit 1-25); profile_get is simpler but both lack documented return schemas. Descriptions are adequate (120-180 chars) but could better explain when to use each tool vs alternatives. No tool annotations present. Error responses return raw HTTP status + body text without actionable recovery steps.
Fetch a public Prompthaus literacy record and career evidence by slug. Use recruiterUrl from search results for hiring review.
Search Prompthaus verified public talent by role or pathway across 18 career pathways (PM, engineering, GTM, finance, energy, defense, markets, life sciences / genomics / biomedical AI, and more). Returns ranked slugs with recruiterUrl and fitSummary. Never invent slugs; only use slugs from this response.
Output schemas not documented. Handlers return generic {content: [{type: 'text', text: string}], isError?: boolean} but actual API response structure (fields in talent_search results, profile_get record format) is never specified. LLMs cannot plan downstream calls or extract structured data.
Error responses lack recovery guidance. Handlers return 'Talent search failed (404): <body>' with no actionable next step. Per pattern:recovery-guide, errors should suggest alternatives: 'No results found. Try broadening the role description or omitting the pathway filter.'
talent_search description mentions '18 career pathways' but does not enumerate them or explain how to discover valid pathway IDs. Parameter description says 'e.g. product-manager, ai-engineer, gtm-growth-engineer, finance-analyst' but does not state these are examples only or how to list all valid values.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 65 | 2026-07-28+ | v2 |
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). Both tools are read-only and idempotent, but this is not declared in the schema. Agents cannot infer safety properties without explicit hints.
profile_get description warns 'never invent slugs' but does not explain what happens if an invalid slug is passed. Does the API return 404? Empty profile? Error message should clarify and suggest calling talent_search first if slug is unknown.