Loading ToolBench
Exposes ModelService over the Model Context Protocol (JSON-RPC 2.0 over stdio) for AI agents
Valem MCP server has 8 tools with basic naming (verb_noun pattern) and descriptions present. However, critical gaps undermine quality: parameter descriptions are minimal (1-2 words), output schemas are completely undocumented, and error handling guidance is absent. Tool names are clear but descriptions lack context on WHEN to use each tool or what downstream operations they enable. No evidence of pagination, result limits, or chaining IDs in responses. Security considerations (permission gates, audit trails) are not visible in the provided code. The server follows a reasonable CRUD pattern but lacks the LLM-optimization details that distinguish production-grade tools.
Create a new reactive computation model with a JSONata expression and optional JSON schema validation
Delete a model by identifier
Evaluate a model with input data and return computed output
Retrieve a model by identifier
Retrieve the current state snapshot of a model
List all available models
Set the state of a model
Update an existing model's expression and/or schema
Output schemas completely undocumented. No visible documentation of what fields each tool returns, forcing LLMs to guess response structure and breaking tool chaining.
Parameter descriptions are minimal (1-2 words: 'Model identifier', 'Input data'). LLMs cannot infer constraints, formats, or valid ranges. Descriptions should be 50-150 chars explaining WHAT, WHEN, and FORMAT.
No error handling guidance. Tools lack recovery hints (e.g., 'Model not found. Call list_models() to see available models'). Agents cannot self-correct on failures.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 48 | 2026-07-28+ | v2 |
list_models lacks pagination parameters (limit, offset, cursor). No indication of result limits or how to handle large model collections. Violates paginated-result pattern.
delete_model and update_model lack confirmation/dry-run support. Destructive operations should offer confirmation step to prevent accidental data loss.