Three tools with complete schemas and descriptions, but significant gaps in parameter documentation and error guidance. lint_markdown and fix_markdown have clear names and basic descriptions (60-80 chars), but lack parameter-level detail on constraints, formats, and error recovery. get_configuration has no parameters but minimal context on when to call it. No output schemas documented. Error handling present but generic (McpError wrapping). No tool annotations (readOnlyHint/destructiveHint). Naming follows verb_noun pattern correctly. Overall: functional but below production baseline for agent-facing tools.

• Add explicit output schemas to each tool. E.g., lint_markdown should return {issues: [{lineNumber, ruleNames, ruleDescription, fixInfo}], filePath, totalCount}. Document field types and meanings.
• Enhance filePath parameter description: 'Absolute or relative path to a .md file. Must exist and be readable. Returns error if file not found or unreadable.'
• Add tool annotations in ListToolsRequestSchema: lint_markdown gets readOnlyHint=true; fix_markdown gets destructiveHint=true, idempotentHint=false (since repeated fixes may differ).
• Improve error messages with recovery guidance. E.g., 'File not found: /path/to/file.md. Check the path and try again.' or 'Permission denied reading file. Ensure the file is readable by the current user.'
• Document get_configuration output: what rules are returned, how to interpret configuration, and when to call it (e.g., 'Call before lint_markdown to understand which rules are active').

Inferred effective spec: <=2025-11-25.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry