A Rust-based MCP server providing tools for multiple external services including arXiv, Discord, HuggingFace, Replicate, Shopify, and Twitter
Critical gaps across all dimensions. Of 39 tools, 9 lack descriptions entirely (PostDmTool, PostMessageTool, PostWebhookTool, SearchModelsTool, GetModelInfoTool, GetModelSampleImagesTool, GetReadmeTool, WhoamiTool across multiple servers). Input schemas are visible for only ~15 tools; the remaining 24 tools have no visible schema definitions in the provided source. Parameter descriptions are sparse or missing. No output schemas documented. Error handling is absent from visible code. Tool names lack consistent verb-noun structure (e.g., 'PostDmTool' vs 'PostDm'). No evidence of idempotency, rate limiting, or secret injection patterns. The codebase registers tools but does not expose their full definitions in the provided excerpts, making comprehensive evaluation impossible.
Add a reaction to a message in a Discord channel.
Assign a role to a Discord user.
24 of 39 tools (61%) have no visible description in source code. Examples: PostDmTool, SearchModelsTool, CreateOrderTool, PostTweetTool.
Input schemas not visible in provided source for 24 tools. Tool definitions are registered but schema details are not shown in excerpts. Cannot verify parameter types, descriptions, or constraints.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 34 | <=2025-11-25 | v2 |
Extract text from an arXiv paper PDF.
Get messages from a Discord channel.
Fetch a specific paper by its ArXiv ID.
Bulk harvest metadata from ArXiv using OAI-PMH interface.
Search for papers by a specific author on ArXiv.
Search for papers on ArXiv using various criteria.
No output schemas documented for any tool. LLMs cannot infer what fields to expect in responses, breaking downstream tool chaining and forcing agents to guess field names.
No error handling guidance visible. Tools do not document recovery paths, retryability, or actionable error messages. Agents cannot determine whether to retry, ask user, or abort.
Destructive tools (DeleteOrderTool, DeleteProductTool) lack confirmation or dry-run patterns. No evidence of permission gates or audit logging. Agents can delete data without safeguards.