MCP server for interacting with the NEAR AI Marketplace to post jobs, check job status, and approve work completions
Three tools with basic schemas and minimal descriptions. Schemas present but sparse: no enums, no constraints, no output documentation. Naming is verb-forward (post_, check_, approve_) but descriptions are 1-2 sentences with no guidance on when to use each tool or what they return. No error handling guidance. approve_completion is irreversible but description does not warn of consequences. Overall: functional but well below production baseline (median ~45-55 for community servers; this lands at lower end due to missing param docs and output schemas).
Accept submitted work and release escrowed NEAR
Check the status of a specific job
Post a new job on the NEAR AI Market
All parameters lack descriptions. Rubric requires every parameter to have a description explaining what it controls. 'job_id', 'title', 'budget_amount', 'deadline_seconds' are undefined for the LLM.
No output schemas documented. LLMs cannot plan downstream calls or extract data without knowing what fields are returned. Rubric baseline: 100% of A+ tools have documented return types.
approve_completion is irreversible (releases escrowed funds) but description lacks warning or confirmation step. Rubric: 'Irreversible operations should support dry-run or confirmation.' No error guidance if job_id is invalid.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 44 | 2026-07-28+ | v2 |
No input validation or error handling guidance in code. If API returns 404 or 400, response is raw JSON with no recovery hint. Rubric: 'Error responses must tell the LLM what to do next.'
post_job has optional parameters (tags, budget_amount, deadline_seconds) with no constraints. No enum for tags, no min/max for budget_amount or deadline_seconds. LLMs will pass arbitrary values.