Server has 24 tools with basic schemas and descriptions, but significant gaps limit production readiness. All tools have descriptions (10-50 chars, mostly generic), and input schemas are present with type definitions. However, descriptions are uniformly brief and lack LLM-optimized guidance on WHEN to use each tool or what to do on failure. Parameter descriptions exist but are minimal (e.g., 'The issue key' without format hints). Output schemas are not documented, LLMs cannot predict response structure for chaining. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear risk levels. Error handling is absent from tool definitions. Naming is consistent (verb_noun pattern) and clear, but composition issues exist: search_issues and get_issue both retrieve issues; analyze_issue is vague about what 'analysis' means. No pagination guidance in list_* tools despite Jira's large datasets.

• Add output schema documentation to every tool. Example: 'Returns: {issue_key: string, summary: string, status: string, assignee: {name: string, email: string}, created: ISO8601, updated: ISO8601}'. This enables LLM chaining.
• Expand tool descriptions to 100-200 chars with WHEN/WHY context. E.g., 'Get full details of a specific issue by key. Use this after search_issues to fetch rich metadata (description, comments, history). Returns all fields including custom fields.'
• Add tool annotations: mark delete_issue and delete_comment with destructiveHint=true; mark all read-only tools with readOnlyHint=true; mark create_issue with idempotentHint=false (retries risk duplicates).
• Document pagination limits in list_* tools: 'maxResults: max 50 per request (default 20). Returns total count. Use startAt to iterate.' Cap default to 20 to prevent context explosion.
• Replace free-form 'analysisType' parameter with enum: ['summary', 'risk_assessment', 'effort_estimate', 'dependency_analysis'] and document each in the description.

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

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