Loading ToolBench
Model Context Protocol (MCP) endpoint exposing a code graph and admin store configuration to AI agents for Magento projects
Server has 5 tools with basic naming and descriptions, but critical gaps in schema documentation and parameter descriptions. Tool names follow verb_noun convention (get_*, search_*), which is good. However, input schemas are either empty objects or minimally documented. Most parameter descriptions are present but generic. No output schemas documented. Error handling and recovery guidance absent. Descriptions average ~80 chars, below the 194-char baseline for A+ tools. No tool annotations (readOnlyHint, destructiveHint) despite all tools being read-only.
Get the schema of the code graph
Get detailed results from a graph search
Get the current status of the Magentic index
Search the code graph for structural code questions (classes, methods, dependencies, plugins, observers, Composer packages)
Search admin store configuration for configuration questions
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields without knowing response structure.
Input schemas are minimal or empty. get_status and get_graph_schema have empty input objects ({}). graph_search and store_config_search have single string params with minimal constraint documentation.
No tool annotations present. All 5 tools are read-only but lack readOnlyHint annotation. This prevents clients from optimizing caching and safety policies.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 48 | 2026-07-28+ | v2 |
No error handling or recovery guidance. Tools lack descriptions of failure modes, retryability, or what the LLM should do if a search returns no results.
Search tools (graph_search, store_config_search) lack pagination parameters (limit, offset, page). Large result sets will blow context windows without pagination support.