Tool: search_screen_content
Parameters:
  query: "search keyword"
  start_time: "2026-02-25T00:00:00Z"  # optional, time range
  end_time: "2026-02-26T00:00:00Z"
  app_name: "Chrome"  # optional, limit to specific app
  limit: 10

Tool: search_audio_transcripts
Parameters:
  query: "meeting discussion content"
  start_time: "2026-02-25T09:00:00Z"
  limit: 5

Tool: get_app_usage
Parameters:
  start_time: "2026-02-25T00:00:00Z"
  end_time: "2026-02-26T00:00:00Z"

# Search screen OCR content
curl "http://localhost:3030/search?q=keyword&content_type=ocr&limit=10"

# Search audio transcripts
curl "http://localhost:3030/search?q=keyword&content_type=audio&limit=5"

# Get recent activities
curl "http://localhost:3030/search?limit=20&start_time=2026-02-25T00:00:00Z"

User: I saw a very good Python library yesterday afternoon, with pipe in its name
→ Search OCR content, limit time to yesterday afternoon
→ Return matching screenshots and context
→ Tell the user: You browsed the xxx project on GitHub in Chrome (14:32)

User: What was discussed in this morning's meeting?
→ Search audio transcripts, limit time to this morning
→ Extract key discussion points and action items
→ Output structured meeting summary

User: How much time did I spend on each app today?
→ Get app usage statistics
→ Generate time distribution report (can be combined with chart generation Skills to generate charts)