Symora provides 8 well-named, read-only code intelligence tools with consistent verb-noun naming (get_, search_, list_, find_, analyze_). All tools have descriptions (avg ~80 chars) and documented input schemas with type definitions. However, output schemas are not documented in the visible code, parameter descriptions are minimal (often single phrases), and error handling guidance is absent. No tool annotations (readOnlyHint, etc.) despite all being read-only. Schemas are present but sparse, most parameters lack detailed constraints, ranges, or format guidance. This is solid foundational work but lacks the LLM-optimization depth of A-grade tools.

• Document output schemas for all 8 tools. Specify returned fields (e.g., search_symbols returns [{symbol_name, file_path, line_number, symbol_type, context}]). Include field types and whether they are always present.
• Expand parameter descriptions to 50 - 150 chars with format guidance. E.g., search_symbols.limit: 'Maximum results to return (1 - 100, default 20). Larger limits may slow search.' get_file_content.start_line: 'Starting line number (1-indexed, inclusive). Omit to start from line 1.'
• Add readOnlyHint: true to all 8 tool definitions in the MCP schema. This signals to agents that these tools are safe to call without confirmation.
• Add error recovery guidance to tool descriptions. E.g., get_symbol_details: 'Returns symbol definition, type, and location. If symbol not found, try search_symbols() with a partial name or broader pattern.'
• Document pagination behavior for search_symbols, list_files, and search_content. Specify result limit (e.g., 'Returns up to 50 results. Use limit parameter to adjust (1 - 100).') and whether a next_cursor or offset is provided for continuation.

Inferred effective spec: 2025-06-18+.

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