Four tools with clear verb-noun naming (search_apps, get_app_info, list_splits, download_apk). Descriptions present and reasonably detailed (60-150 chars each). Input schemas visible with type definitions and descriptions for all parameters. However, output schemas are NOT documented, LLMs cannot predict response structure. Error handling is minimal; no recovery guidance or actionable error messages. Missing tool annotations (readOnlyHint/destructiveHint). Parameter descriptions lack format constraints and ranges. No pagination support despite search_apps returning lists.

• Document output schemas for all four tools. Show the exact JSON structure LLMs will receive (e.g., search_apps returns [{package, title, developer, rating}], get_app_info returns {title, developer, version, rating, url}).
• Add pagination to search_apps: include offset/limit in request, return total_count and next_offset in response. Cap results at 20-50 items per call.
• Add tool annotations via mcp-go: mark search_apps and get_app_info with readOnlyHint=true, mark download_apk with destructiveHint=true.
• Enhance error messages with recovery guidance. E.g., 'App "com.invalid" not found. Try search_apps("app name") to find the correct package.'
• Add parameter constraints to descriptions: 'limit: integer, 1-100 (default 10)', 'version_code: positive integer or omit for latest', 'output_dir: valid filesystem path, must be writable'.