MCP server to share professional information with LLMs
Server provides 5 tools with basic schemas and descriptions. All tools are read-only (good for safety). However, definitions lack critical elements: output schemas are undocumented, parameter descriptions are minimal, error handling is absent, and many parameters lack type information or descriptions. Naming is generally correct (verb-first convention), but descriptions are too brief to guide LLM tool selection effectively. No evidence of pagination support despite potential for large result sets. Tool composition is adequate but lacks context-chaining IDs for multi-step workflows.
Obtiene detalles específicos de una experiencia laboral
Obtiene información del perfil profesional por sección
Obtiene detalles específicos de un proyecto
Obtiene habilidades filtradas por categoría
Busca información específica en el perfil profesional
Output schemas are completely undocumented. LLMs cannot plan downstream tool calls or extract required fields without knowing response structure.
Parameter descriptions are too brief (5-30 chars). 'ID de la experiencia laboral' and 'ID del proyecto' lack context on format, where to find IDs, or how to obtain them. LLMs cannot disambiguate.
No error handling or recovery guidance. Tools will fail silently on invalid IDs (e.g., experienceId='invalid') with no actionable message for the LLM.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 46 | 2026-07-28+ | v2 |
Tool descriptions are 35-50 chars, below baseline (194 chars). Too terse to convey WHAT, WHEN, and WHY. E.g., 'Busca información específica en el perfil profesional' does not explain when to call search_profile vs get_profile.
No pagination support declared or implemented. If profiles have many experiences or projects, results could exceed context limits without offset/limit parameters.
Parameter 'experienceId' and 'projectId' are opaque system IDs. No guidance on format or how to discover valid IDs. Should accept human-friendly names (project name, experience title) or explain where to get IDs.
Tool descriptions are in Spanish; MCP protocol and LLM communication typically expects English. Mixed-language server reduces compatibility and confuses LLM parsing of capabilities.
No evidence of response structure in tool handlers. Handler code (line: src/tools/handlers) is not shown, so output schema and field types cannot be verified. Assumed to return unstructured text.