Two simple, focused tools with clear names and basic schemas. Both tools have descriptions and input schemas with type definitions. However, parameter descriptions are minimal or missing (e.g., 'bytes' param in format tool lacks description; 'input' param in parse tool lacks description). Output schemas are not documented. Error handling is present but generic ('bytes failed: ...'). No tool annotations (readOnlyHint, etc.). Naming is verb-based and clear (format, parse), but descriptions could be more detailed about when to use each tool and what they return.

• Add descriptions to all parameters. E.g., 'bytes': 'The byte count to format (must be a finite number)'; 'input': 'Size string to parse, e.g. "1.5 MB", "512KiB", or a bare number (treated as bytes)'.
• Document output schemas in tool descriptions. E.g., 'Returns an object with a "formatted" field containing the human-readable string' for format; 'Returns an object with a "bytes" field containing the parsed byte count' for parse.
• Expand tool descriptions to explain when to use each tool and the difference between binary (1024-based) and SI (1000-based) units. E.g., 'Use binary for storage (KiB/MiB); use SI for network speeds (kB/MB).'
• Add tool annotations to the inputSchema. Mark both tools with 'readOnlyHint: true' and 'idempotentHint: true' to signal to the LLM that they are safe to call repeatedly.
• Improve error messages to guide recovery. E.g., 'Invalid bytes value: must be a finite number (got NaN)' or 'Unknown unit in "512 XB": valid units are B, KiB, MiB, GiB, TiB, PiB, EiB, kB, MB, GB, TB, PB, EB, K, M, G, T'.