A collection of MCP (Model Context Protocol) servers including weather information, content design assistance, documentation helpers, and meeting notes organization
Repository contains 14 tools across 4 servers (weather, content-design, documentation, meeting-notes). Tool definitions are visible in source but lack rigor. Descriptions range from adequate (analyze_readability: 156 chars) to minimal (weather_inquiry: 67 chars). Input schemas are present but sparse, most tools accept 1-2 string parameters with minimal type information. No output schemas documented. Error handling is absent from visible code. Tool naming follows verb_noun convention (get_weather, search_documentation) but several tools lack clear action semantics (weather_inquiry, weather_travel_advice read as templates, not executable tools). No evidence of parameter validation, enum constraints, or recovery guidance. Composition is reasonable (separate tools for search vs summarize) but chaining support unclear, no documented output field mappings to downstream tool inputs.
Analyze text readability and provide metrics and recommendations. Checks sentence length, word complexity, paragraph structure, and readability score estimation.
Check content for accessibility issues and provide recommendations. Checks for vague link text, missing context, all-caps text, special characters that may confuse screen readers, and abbreviations without expansion.
Evaluate content tone and check if it matches the desired tone. Analyzes formality level, emotional tone, personality indicators, and consistency.
Extract action items from meeting transcript or notes. Identifies tasks mentioned, assigned owners (if mentioned), due dates (if mentioned), and priority level.
Extract questions raised during the meeting (answered and unanswered). Identifies questions asked, whether they were answered, and follow-up needed.
Find code examples across documentation files. Searches for code blocks, inline code, and specific programming language examples.
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract structured data without knowing response field names and types.
weather_inquiry and weather_travel_advice lack clear action semantics. Descriptions label them as 'templates' rather than executable tools, creating ambiguity about when/how to invoke them.
Input schemas are minimal, most tools accept only string parameters with no enum constraints, format patterns, or validation rules. LLMs cannot self-correct invalid inputs.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 46 | 2026-07-28+ | v2 |
Get current weather information for a specified city. Returns temperature data for the requested location.
Identify key decisions made during the meeting. Looks for agreement statements, conclusions reached, approved plans, and rejected options.
Review error messages for clarity, helpfulness, and actionability
Search for content across documentation files. Searches for keywords and phrases, function/class names, concepts and topics, and code examples.
Summarize main discussion points from the meeting. Creates a concise summary of main topics discussed, key points made, and important context.
Summarize a documentation file's content and structure. Provides overview of content, main sections/headings, key concepts covered, and code examples present.
Template for asking about weather conditions in a specific location
Template for getting weather-based travel advice for a destination
No error handling guidance visible in tool definitions. Tools provide no recovery hints, categorization of error types, or actionable next steps for LLMs.
No tool chaining support documented. Output fields from one tool (e.g., search_documentation) do not explicitly map to input parameters of downstream tools (e.g., summarize_documentation).