feat: Add tool annotations for improved LLM tool understanding

[bryankthompson]

Summary

Adds MCP tool annotations (readOnlyHint, destructiveHint, title) to all 15 tools to help LLMs better understand tool behavior and make safer decisions about tool execution.

Changes

• Added readOnlyHint: true to 10 read-only tools:

  • plan_task - Generates planning prompts
  • analyze_task - Analyzes task complexity
  • reflect_task - Reviews completed tasks
  • list_tasks - Lists all tasks
  • verify_task - Verifies task completion
  • query_task - Queries tasks
  • get_task_detail - Gets task details
  • process_thought - Processes thought chains
  • init_project_rules - Generates project rules
  • research_mode - Generates research prompts

• Added destructiveHint: true to 5 tools that modify data:

  • split_tasks - Creates subtasks in database
  • execute_task - Updates task status to in_progress
  • delete_task - Deletes tasks from database
  • clear_all_tasks - Clears all tasks (irreversible)
  • update_task - Updates task content

• Added title annotations for human-readable display

Why This Matters

• Annotations provide semantic metadata that helps MCP clients understand tool behavior
• Clients like Claude Code can auto-approve tools with readOnlyHint: true without user confirmation
• Destructive tools trigger confirmation prompts for safer operation
• Enables smarter UI/UX decisions based on tool characteristics

Testing

• Server builds successfully (npm run build)
• Live verification: Started server and confirmed tools/list returns annotations for all 15 tools
• Annotation values match actual tool behavior (read-only vs destructive)

Before/After

Before:

{
  name: "delete_task",
  description: "...",
  inputSchema: {...}
  // No annotations
}

After:

{
  name: "delete_task",
  description: "...",
  inputSchema: {...},
  annotations: {
    title: "Delete Task",
    destructiveHint: true
  }
}

---

🤖 Generated with Claude Code