MCP server for job hunting automation, providing tools for job scraping, candidate expectations analysis, commute cost calculation, company reputation scoring, and financial viability assessment.
The JobHunter server exposes only one tool: 'get_commute_cost'. While the tool has a name and basic description, it lacks critical schema documentation, parameter descriptions, and output specification. The description is generic and does not follow LLM-optimized conventions. The tool definition appears inferred rather than explicitly registered with proper MCP tooling. The source code shows the tool is referenced in scripts/process_jobs.py but no explicit tool registration, input validation, or output schema is visible in the codebase.
Calculate commute costs and overnight requirements for job destinations
No input parameter descriptions visible in source code. Parameters 'destination' and 'origin' have minimal descriptions ('City or destination for commute calculation', 'Starting location for commute') that do not explain expected format, constraints, or valid values.
No output schema documented. The tool description does not state what fields are returned, what data types to expect, or how to use the result. LLMs cannot plan downstream operations without knowing the response structure.
Tool description is generic and lacks LLM-actionable guidance. 'Calculate commute costs and overnight requirements for job destinations' (63 chars) does not explain when to call this tool, what context it requires, or what use case it solves. Does not clarify whether it returns costs in a specific currency, overnight stay hours, accommodation recommendations, or all of these.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 38 | 2026-07-28+ | v2 |
Parameters lack type definitions. The source shows input parameters as strings ('destination', 'origin') but no JSON Schema types, formats, constraints, or validation rules are visible.
No error handling guidance. Tool description does not explain what happens if destination is invalid, unreachable, or in an unsupported country. No recovery guidance for LLMs (e.g., 'Try search_cities() first' or 'Destination must be a known city name').
Tool definition appears inferred rather than explicitly registered. The .mcp.json shows an HTTP endpoint ('https://mcp400.debytes.io/mcp') but no MCP server code is visible that registers tools with proper schemas.
No parameter constraints or enums. 'destination' and 'origin' accept free-form strings with no validation. Should define: accepted city formats, supported countries, length limits, or enum of known cities to prevent hallucinated values.
No pagination or result limiting. If commute cost results can be large or multi-step, no indication of how results are structured, paginated, or limited. Missing limit/offset parameters or result count.