Pagestress MCP server for read-mostly governed run, artifact, and review workflows.
Critical gaps across all dimensions. Tool names lack action verbs (uiq_api_* prefix is domain-specific but not action-oriented). Descriptions are minimal API endpoint strings (10-50 chars), far below the 194-char baseline for production tools. Input schemas present but severely underdocumented: most parameters lack descriptions or have trivial ones. No output schemas documented. Error handling absent. The 'uiq_api_automation_run' tool explicitly warns that 'env' is not accepted in strict mode, but this constraint is buried in the description rather than enforced via schema validation or clear parameter documentation. Tools 6-8 use generic 'action' enums (list|get|create|update) bundling multiple operations into single tools, violating single-responsibility principle. No evidence of idempotency, pagination limits, or recovery guidance.
POST /api/automation/tasks/{taskId}/cancel
GET /api/automation/commands
POST /api/automation/run (strict mode: params only)
GET /api/automation/tasks/{taskId}
GET /api/automation/tasks
Universal flows API: list|get|import_latest|create|update
Universal runs API: list|get|create|otp|cancel
Tool names lack action verbs. All 8 tools use 'uiq_api_*' prefix with no clear verb (get, create, update, delete, list). LLMs cannot infer intent from names alone. Example: 'uiq_api_flows' could mean list, get, create, or update flows.
Descriptions are API endpoint strings (10-50 chars), far below 194-char baseline. No WHAT/WHEN/WHY context. Example: 'GET /api/automation/commands' tells LLM nothing about when to call it or what it returns.
Multi-operation tools (uiq_api_flows, uiq_api_templates, uiq_api_runs) bundle list|get|create|update into single tools via 'action' enum. Violates single-responsibility principle. LLM must reason about which action to pick; tool composition is opaque.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 40 | 2025-06-18+ | v2 |
Universal templates API: list|get|export|create|update
No output schemas documented. LLMs cannot plan downstream calls or extract required fields. No pagination guidance (limit/offset/cursor). No indication of result structure or field types.
Parameter descriptions missing or trivial. Example: 'status' param in uiq_api_automation_tasks has description 'Filter tasks by status' but no enum values, format, or valid options. LLM must guess valid values.