Enterprise MCP server with ServiceNow integration and health monitoring
Enterprise MCP has 9 tools across health monitoring and ServiceNow integration. Naming is generally clear and action-oriented (get_*, create_*, update_*). However, critical gaps exist: descriptions are present but mostly generic (10-50 chars), parameter descriptions are minimal or missing, input schemas lack depth, and output schemas are entirely undocumented. The server shows basic structure but lacks the detail required for confident LLM tool selection and composition. No visible error handling guidance, no pagination support for list tools, and no tool annotations (readOnlyHint/destructiveHint). The update_incident tool is particularly sparse, it accepts only incident_id with no description of what fields can be updated.
Create a new ServiceNow incident.
Get ServiceNow change requests.
Get CPU usage statistics.
Get a specific ServiceNow incident by ID.
Get ServiceNow incidents.
Get memory usage statistics.
Get comprehensive system health information.
No output schemas documented. Callers and downstream tools cannot know what fields to expect from any tool response, breaking tool composition and forcing LLMs to infer structure.
update_incident has minimal schema: only incident_id parameter with no description, and no indication of which fields can be updated (status, priority, assignment, description, etc.). LLMs cannot infer what this tool modifies.
Tool descriptions are generic and under 50 characters for most health tools. 'Get CPU usage statistics' does not explain when to call this vs get_memory_usage, what it returns, or whether to use it for alerting decisions.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 49 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 42 | - | v1 |
Get server uptime information.
Update a ServiceNow incident.
No pagination parameters on list tools (get_incidents, get_change_requests). Default limit=10 is provided but no offset/cursor or total_count in response schema, blocking iteration over large result sets.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). create_incident and update_incident should be marked destructiveHint=true or modifyingHint=true so agents know these have side effects and may need confirmation.
Parameter descriptions missing or minimal. 'limit' on get_incidents has a description, but 'short_description', 'description', 'category' on create_incident lack format/length constraints. 'priority' defaults to 3 but no description of 1-5 scale or meaning.
No error handling guidance visible. Tools do not document what errors are retryable, which require user input, or what recovery actions to take if ServiceNow API fails.