MCP server for DevChain local app providing tools for agent management, chat messaging, document operations, and epic/status management
DevChain MCP server has 7 tools with basic schemas and descriptions, but significant gaps prevent production readiness. All tools have descriptions (10-100 chars, mostly adequate) and input schemas with type definitions. However, output schemas are completely undocumented, LLMs cannot predict response structure for chaining. Parameter descriptions are present but often generic ('Optional search query', 'Optional tags'). No error handling guidance, no recovery paths, and no tool annotations (readOnlyHint/destructiveHint). Naming is verb-first and clear (list_, get_, send_, create_, update_), following conventions well. Critical issue: sessionId appears in 5 tools but is never explained, what does it resolve? How does it work? This forces LLMs to guess. Transport is unknown (likely STDIO based on Dockerfile context), which caps protocol readiness severely.
Create a new document in a project
Get detailed information about a specific agent including profile, teams, presence, and assigned epics
Get detailed information about a specific document with optional link resolution
List agents and guests in a project with pagination, search, and presence information
List documents in a project with optional filtering by tags and search query
Send a message to agents or teams within a project or to another project
Update an existing document
No output schemas documented. LLMs cannot predict response structure, breaking tool chaining. E.g., list_agents returns what fields? Does it include team_id for downstream send_message calls?
sessionId parameter appears in 5 tools but is never explained. What does it resolve? Is it a user session, project context, or auth token? Undocumented dependencies force LLMs to guess.
No error handling guidance. Tools lack recovery hints (e.g., 'If agent not found, try list_agents() first'). Errors will not guide LLM to next action.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 53 | <=2025-11-25 | v2 |
No tool annotations. send_message and update_document are destructive/write operations but lack destructiveHint. create_document and list_agents lack idempotentHint. LLMs cannot reason about retry safety.
Parameter descriptions are generic. 'Optional search query' does not explain what fields are searched (name only? description? tags?). 'Optional tags' does not explain tag format or valid values.