Loading ToolBench
MCP server for analyzing React projects, providing FSD (Feature-Sliced Design) violation detection, naming convention checking, project scaffolding, and rule management
This MCP server defines 3 tools for FSD architecture linting and React entity scaffolding. Tool names follow verb_noun convention (analyze_project, get_rules, scaffold_entity), which is good. However, schema completeness is poor: the source code shows only partial input schemas for analyze_project and get_rules (parameters visible but incomplete in the truncated code snippet), and the full schema for scaffold_entity is not visible in the provided source. Descriptions are present but generic (10-50 chars), leaving LLMs without sufficient context to disambiguate tool selection. No documented output schemas are visible. Error handling guidance is absent. The server appears to be a Next.js frontend with MCP integration via STDIO, not a production-grade remote MCP server.
Analyze project files for FSD violations and naming conventions.
Get current FSD and naming project rules.
Generate a standardized FSD entity structure with API, types, schema, and converters.
Descriptions are extremely brief (10-50 characters) and lack LLM-optimized context. According to baseline rubric, good descriptions are 50-200 chars and explain WHAT, WHEN, and WHY. These descriptions ('Analyze project files for FSD violations...', 'Get current FSD and naming project rules.', 'Generate a standardized FSD entity structure...') are too terse and do not guide LLM tool selection when multiple tools are available.
Input schemas are incomplete in the source code provided. analyze_project shows 'files' (array of strings) and 'codeSnippets' (optional array of objects), but the truncated output does not show full schema details. get_rules shows 'domain' (enum) and 'pattern' (string, optional) but enum values are partially visible ('fsd','naming','linter','structure','styles','patterns'). scaffold_entity schema is not visible at all. Without complete schema definitions in the source, cannot verify all parameter types and descriptions are present.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 43 | 2026-07-28+ | v2 |
No output schemas documented. The rubric baseline states 100% of A+ tools have documented return types. This server provides no visible schema for what analyze_project, get_rules, or scaffold_entity returns. LLMs cannot plan downstream tool calls or extract the right data without knowing output structure.
Parameter descriptions are missing or minimal. The rubric requires every input parameter to have a non-empty description. In the visible schema for analyze_project, 'codeSnippets' has a description, but incomplete visibility into scaffold_entity and get_rules parameters makes full assessment difficult. Where visible, descriptions exist but are generic.
No error handling guidance. The server provides no examples of error responses or recovery hints. According to the rubric, error responses must tell the LLM what to do next. A user calling analyze_project with invalid file paths should get 'File not found. Check that paths are relative to project root.' Not visible in the source.
Missing state-change declarations. scaffold_entity has a WRITE risk (modifies filesystem), but the tool description does not explicitly state 'This tool creates files on disk' or indicate it is irreversible. Agents need to know which calls have side effects. analyze_project and get_rules are READ_ONLY (good) but this is not declared in their descriptions.