Server has 5 tools with basic schemas and descriptions. All tools have input schemas with typed parameters and descriptions present. However, descriptions are generic (avg ~80 chars, below the 194-char baseline for A+ tools), lack WHEN/WHY context, and omit error recovery guidance. Parameter descriptions are minimal (e.g., 'Repository owner' vs 'The GitHub username or organization name that owns the repository'). No output schemas documented. Missing idempotency hints on write operations (triage_issue, add_comment). No pagination support on list_issues despite returning up to 10 items. Error handling is absent, no guidance on what to do if a GitHub API call fails.

• Expand tool descriptions to 100-200 chars. Include WHEN to use (e.g., 'Call this to see open issues before triaging') and WHAT it returns (e.g., 'Returns issue number, title, and current labels'). Reference pattern:tool-description.
• Document output schemas for all tools. E.g., list_issues returns {issues: string, count: number}. This enables LLMs to plan multi-step workflows.
• Add idempotentHint and destructiveHint annotations to triage_issue and add_comment. Mark triage_issue as destructive (modifies state) and add_comment as non-idempotent (repeated calls create duplicates).
• Implement error handling with recovery guidance. Catch octokit errors and return structured responses like {error: 'User not found', suggestion: 'Try search_users() with a partial name'}. Reference pattern:recovery-guide.
• Add pagination to list_issues: accept limit (1-100, default 20) and page (default 1) parameters. Return total_count and has_more fields. Reference pattern:paginated-result.