Single tool with minimal definition quality. Tool name 'general_search' is generic and lacks action clarity (should be 'search_repositories'). Description is adequate (56 chars) but lacks context on when to use vs alternatives, prerequisites, or return structure. Input schema present but parameter description is minimal (27 chars, below 72-char baseline). No output schema documented, LLM cannot predict response structure. No error handling guidance. No pagination support despite search results potentially being large. Tool follows basic FastMCP registration but lacks production-grade polish.

• Rename 'general_search' to 'search_repositories' to match verb_noun convention and clarify intent.
• Document output schema: {total_count: int, items: [{name: string, url: string, description: string|null}]}. Include in docstring.
• Add pagination: limit (1-100, default 20) and offset (default 0) parameters. Return total_count to enable agent-driven pagination.
• Expand parameter description: 'Search term for GitHub repositories (e.g., "python machine learning"). Supports GitHub search syntax (language:python, stars:>1000).'
• Add error handling: catch rate_limit_exceeded, invalid_token, and malformed_query; return structured errors with retry guidance and available alternatives.