CLI toolkit for MCP server contract management: dumps capabilities, generates changelogs and renders documentation
This is a CLI toolkit, not an MCP server. It does not implement the Model Context Protocol, it is a standalone command-line tool for analyzing MCP servers. Tool definitions are inferred from command names in src/index.ts (dump, split, convert, document, diff, breaking, changelog, compare, validate, rules, completion, agents) rather than explicitly registered via MCP's tool mechanism. Only 2 of 12 tools have visible input schemas (dump, diff, breaking, changelog, agents); the rest lack documented parameters entirely. Descriptions are present but generic (e.g., 'Split large MCP descriptions into focused subsets' for split). No output schemas are documented. Error handling is not visible in the provided code. This is fundamentally a CLI application, not an MCP server, and cannot be evaluated as one.
AI coding assistant reference guide - provides markdown-formatted documentation optimized for AI coding assistants
Spot non-backward compatible changes from a structural diff
Generate human-readable changelog from analysis JSON
Human-readable comparison report in one step (diff + breaking + changelog)
Generate shell completion scripts (bash, zsh, fish)
Convert MCP descriptions between formats
Compare two versions and generate structural diff
Not an MCP server. This is a CLI toolkit. Tool definitions are inferred from command names, not explicitly registered via MCP protocol. Cannot be used as an MCP tool by hosted clients or agents.
10 of 12 tools lack documented input schemas. Only dump, diff, breaking, changelog, and agents have visible parameter definitions. split, convert, document, compare, validate, rules, completion have no schema.
No output schemas documented for any tool. Agents cannot infer what fields to expect or plan downstream operations.
Descriptions for 7 tools are under 50 characters and lack context on when/why to use them. E.g., 'Split large MCP descriptions into focused subsets' does not explain the use case or expected input/output.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 37 | 2026-07-28+ | v2 |
Render documentation from MCP descriptions
Extract capabilities from a live MCP server
Browse and explore backward compatibility rules catalog
Split large MCP descriptions into focused subsets
Validate files against JSON schemas (mcpdesc, dump, diff)
No error handling guidance visible. Tools do not document what errors can occur, whether they are retryable, or what the agent should do next.