All three tools have critical definition gaps. Tool descriptions are generic boilerplate ('This is a tool from the github MCP server') that provide no actionable context for LLM selection. The CHECK_RESUME_TOOL has an empty inputSchema with no properties, violating the parameter documentation requirement. ANALYZE_CODEBASE_TOOL and ENHANCE_RESUME_WITH_PROJECT_TOOL accept optional 'directory' parameters but lack descriptions explaining expected format, constraints, or defaults. No output schemas are documented anywhere. Tool names use verb_noun pattern (good), but descriptions fail to explain WHEN to use each tool, WHAT it returns, or dependencies between them. Error handling in implementation (try/catch blocks) exists but provides no recovery guidance to LLMs.

• Replace boilerplate descriptions with specific, actionable text. Example: 'Analyzes the codebase in the specified directory (or current working directory if omitted) and returns detected technologies, languages, frameworks, and recent commit history. Use this to understand project tech stack before enhancing resume.'
• Add parameter descriptions: 'directory (string, optional): Absolute or relative path to project root. Defaults to current working directory. Must be a valid directory path.'
• Document output schemas for all tools. Example for github_analyze_codebase: 'Returns {analysis: {technologies: string[], languages: string[], frameworks: string[], recentCommits: Commit[]}, summary: string}'
• Rename tools to match actual behavior. Consider 'analyze_local_codebase' and 'enhance_resume_from_project' to clarify they operate on local filesystem, not GitHub API.
• Add error recovery guidance in descriptions. Example: 'If directory does not exist, returns error with available parent directories. If no resume found, creates sample resume from GitHub profile.'

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

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