Server has 12 tools with explicit Zod schemas and descriptions. Naming follows verb_noun convention (get_*, sync_*). However, parameter descriptions are sparse or missing in many tools. Output schemas are not documented in the visible code. Error handling exists but lacks recovery guidance. Tool definitions are properly registered in src/tools/index.ts with zodToJsonSchema conversion, but the actual tool implementations (only ping.ts visible) show minimal parameter documentation. Average tool description length appears adequate (~100-150 chars inferred), but parameter-level descriptions are underdeveloped. No tool annotations (readOnlyHint/destructiveHint) present despite clear risk classifications.

• Add parameter descriptions to every Zod schema. For each field in inputSchema, include a description explaining what it controls, valid range/enum, and format (e.g., 'repo_name (string, 1-100 chars): GitHub repository name, e.g., "my-project"').
• Document output schemas for all tools. Add a comment or JSDoc block showing the structure returned (e.g., '{ repos: [{ id, name, stars, language }], total: number }') so LLMs can plan downstream calls.
• Add tool annotations via MCP SDK. Mark sync_github_data with destructiveHint=true. Mark all read-only tools with readOnlyHint=true. This enables agents to reason about side effects.
• Enhance error messages with recovery guidance. When a repo is not found, return 'Repo "xyz" not found. Try get_all_repos() to see available repositories.' instead of a bare error code.
• Implement pagination for list tools. Add limit (default 20, max 100) and offset/cursor parameters to get_all_repos, get_open_items, and generate_weekly_report. Return total_count and next_cursor in responses.

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

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