MCP server for OpenProject integration, providing tools to create projects, work packages, and manage dependencies in OpenProject
Server has 6 tools with basic schemas and descriptions, but significant gaps limit production readiness. Tool names follow verb_noun convention (good), but descriptions are inconsistent in depth. Most parameters lack detailed constraints (enums, ranges, formats). Output schemas are not documented. Error handling is minimal, no recovery guidance or actionable error messages. Security concerns: no visible input validation, no audit logging, no permission gates. Composition is reasonable (tools are single-purpose), but missing pagination for list operations and no documented chaining IDs.
Create a new project in OpenProject.
Create a work package in a project with dates for Gantt chart.
Create a dependency between two work packages for Gantt chart visualization.
Get all projects from OpenProject.
Get work packages from a project in OpenProject.
Health check tool to verify OpenProject MCP Server is running and connected.
Output schemas not documented. LLMs cannot plan downstream calls or extract required fields. No visibility into what create_project, create_work_package, or get_* tools return.
Parameter descriptions lack constraints. 'start_date' and 'due_date' say 'YYYY-MM-DD format' but no validation or error guidance if LLM passes wrong format. 'relation_type' lists enum values in description but not as formal enum constraint.
No pagination support. get_projects and get_work_packages lack limit/offset/page parameters and total count in response. Large result sets will blow context window.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 58 | <=2025-11-25 | v2 |
No error handling guidance. Tools return success/failure but no recovery hints. If create_project fails due to duplicate name, LLM has no guidance on next steps (retry? search existing? ask user?).
No input validation visible in source. Tool implementations not shown, but no evidence of sanitization against injection, format validation, or range checks. Assumes LLM always passes valid input.