Single tool 'codegraph_search' has minimal definition quality. Description is present (67 chars) but lacks actionable context for LLM selection. Input schema exists but contains only one parameter with minimal description (18 chars). No output schema documented. No error handling guidance. No parameter constraints (enum, range, format). Tool name lacks clarity on what 'codegraph' means to an unfamiliar LLM. No indication of prerequisites, return structure, or when to use this vs. alternatives. Baseline: A+ tools average 194 chars description + 72 chars per param; this tool is 67 + 18 = well below threshold.

• Expand tool description to 150-250 chars: explain that it searches a code graph for symbols/definitions/references, what 'codegraph' is (a code analysis index?), and what results contain (file paths, line numbers, symbol types).
• Expand 'query' parameter description to 80-120 chars: specify format (free text? regex? symbol name?), example queries ('function foo', 'class Bar'), and constraints (max length, special chars allowed).
• Document output schema with typed fields: e.g. { results: [{ file: string, line: number, symbol_name: string, symbol_type: enum(function|class|variable), context: string }], total_count: number }.
• Add error handling guidance: 'If no results found, try a shorter query or check symbol spelling. If query is malformed, use alphanumeric + underscore only.'
• Rename tool to 'search_code_symbols' or 'search_definitions' to improve clarity for LLMs unfamiliar with 'codegraph' domain.

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

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