feat: Add tool annotations for improved LLM tool understanding

[triepod-ai]

Summary

Adds MCP tool annotations (readOnlyHint, destructiveHint, title) to all 5 tools to help LLMs better understand tool behavior and make safer decisions about tool execution.

Changes

Tool	Annotation	Reason
spec-workflow-guide	readOnlyHint: true	Only loads workflow guide text
steering-guide	readOnlyHint: true	Only loads steering guide text
spec-status	readOnlyHint: true	Reads spec status from filesystem
approvals	destructiveHint: true	Creates/deletes approval requests
log-implementation	destructiveHint: true	Creates log entries (writes files)

Why This Matters

• Annotations provide semantic metadata that helps LLMs understand tool behavior
• MCP clients like Claude Code can auto-approve read-only tools without user confirmation
• Destructive tools trigger confirmation prompts for safer execution
• Enables better tool discovery and filtering by characteristics

Testing

• Server builds successfully (npm run build)
• Live verification: Started server and confirmed tools/list returns annotations for all 5 tools
• Annotation values match actual tool behavior

Verification Output

Total tools: 5

Tool Name                 Title                     ReadOnly   Destructive
----------------------------------------------------------------------
spec-workflow-guide       Spec Workflow Guide       True       N/A       
steering-guide            Steering Guide            True       N/A       
spec-status               Spec Status               True       N/A       
approvals                 Approvals                 N/A        True      
log-implementation        Log Implementation        N/A        True

🤖 Generated with Claude Code