Single tool 'codegraph_explore' has a basic schema with three parameters (query, maxFiles, projectPath) but critical gaps undermine quality. The tool description is generic (47 chars) and lacks context on WHEN to use it or WHAT it returns. Parameter descriptions are minimal: 'query' lacks format guidance, 'maxFiles' has no range constraints, 'projectPath' is vague about resolution. No output schema is documented. The tool name 'codegraph_explore' is acceptable (verb_noun form) but the implementation lacks the rigor expected of production tools. No error handling guidance, no pagination support despite being a search tool, no indication of result limits. The schema is visible but incomplete, missing type constraints, ranges, and output documentation.

• Expand tool description to 100-200 chars. Example: 'Search and explore code graph indexes across a project. Returns matching files with line numbers and relevance scores. Use this to understand codebase structure before refactoring or asking follow-up questions about specific files.'
• Document output schema explicitly. Specify: array of objects with fields {file_path, line_number, snippet, relevance_score, language}. Include example response.
• Add range constraints to maxFiles: 'Maximum number of files to return (1-100, default 20). Larger values may timeout.'
• Clarify 'projectPath' parameter: 'Absolute or relative path to project root. If omitted, uses current working directory.'
• Add pagination support: accept 'offset' and 'limit' parameters, return 'total_count' and 'has_more' in response. Document default limit (20) and max limit (100).