Server has 4 tools with basic schemas and descriptions, but significant gaps limit production readiness. Tool names follow verb_noun convention (getIssues, getProjects, getIssue, initialize), which is good. However, descriptions lack LLM-optimization guidance (e.g., 'when to use this vs similar tools'), parameter descriptions are minimal, and output schemas are completely undocumented. The initialize tool exposes API credentials as parameters, violating secret-injection patterns. Error handling returns generic messages without recovery guidance. No tool annotations (readOnlyHint, destructiveHint) despite clear read/write semantics.
Fetches detailed information for a specific issue by ID. In addition to basic fields, it can include related data such as journals, attachments, relations, changesets, watchers, and child issues. Parameters: id, include (for related entities).
Retrieves a list of issues from Redmine. Supports filtering by project, status, assignee, and more, as well as sorting and pagination. Returns basic information for each issue, including ID, subject, description, status, priority, assignee, author, creation date, and update date. Parameters: project_id, status_id, assigned_to_id, tracker_id, sort, limit, offset.
Lists all accessible projects with basic information like ID, name, identifier, description, status, created and updated dates. Parameters: limit, offset.
Initializes the Redmine MCP server with base URL and API key. This tool is only exposed if automatic initialization from environment variables fails.
API credentials (baseUrl, apiKey) exposed as tool parameters in initialize tool. Credentials will be logged in agent traces and prompt history.
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields (e.g., issue_id for chaining). Responses are raw JSON dumps without field descriptions.
Error messages lack recovery guidance. 'Error: 404' or generic 'API connection failed' do not tell the LLM what to do next (retry, ask user, try alternative tool).
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). getIssues/getIssue/getProjects are read-only but not marked; initialize is destructive (modifies server state) but not marked.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 49 | 2026-07-28+ | v2 |
Parameter descriptions are minimal (e.g., 'Filter by project ID' lacks format/range guidance). 'sort' parameter has no enum or valid values documented; LLMs will guess.