The official MCP Server for the Merchant Categories API
Server has 11 tools with basic naming and descriptions, but significant gaps in schema completeness, parameter documentation, and error handling. Most tools lack detailed input schemas and parameter descriptions. The three dynamic tools (list_api_endpoints, get_api_endpoint_schema, invoke_api_endpoint) have partial schemas visible but lack comprehensive parameter documentation. No evidence of output schema documentation, pagination support, or error recovery guidance. Descriptions are present but generic (10-50 chars), below the 50-200 char baseline for LLM-optimized tools.
Create a new category
Delete a category
Get the schema for an endpoint in the Merchant Categories TypeScript API. You can use the schema returned by this tool to invoke an endpoint with the `invoke_api_endpoint` tool.
Invoke an endpoint in the Merchant Categories TypeScript API. Note: use the `list_api_endpoints` tool to get the list of endpoints and `get_api_endpoint_schema` tool to get the schema for an endpoint.
List or search for all endpoints in the Merchant Categories TypeScript API
List all categories
Eight core tools (create/retrieve/update/delete/list_categories, list_children_categories, list_products_categories, search_categories) have NO visible input schemas in source code.
Tool descriptions are generic and under 50 characters (e.g., 'Create a new category', 'Retrieve a category by ID'). Descriptions lack context on WHEN to use each tool, what it returns, or dependencies. Baseline for LLM-optimized tools is 50-200 chars with actionable guidance.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 38 | 2026-07-28+ | v2 |
List child categories of a parent category
List products in a category
Retrieve a category by ID
Search for categories
Update an existing category
No evidence of parameter descriptions for core tools. LLMs cannot infer what 'category_id', 'parent_id', or 'query' parameters mean without explicit descriptions. This violates the requirement that every parameter must have a description.
No documented output schemas for any tool. LLMs cannot plan downstream calls or extract required fields (e.g., category_id for chaining) without knowing what each tool returns. This blocks tool composition.
No error handling guidance visible. Tools lack recovery hints (e.g., 'If category not found, try search_categories()') or error classification (retryable vs. user-fixable). Agents cannot self-correct on failures.