name: sync-data-linkedin description: > Fetch LinkedIn conversations from browser into a JSON file. No classification, no actions — just data. Requires browser. allowed-tools: Read, Write, Edit, Glob, Grep, Bash(python3:), mcp__claude-in-chrome__

Sync LinkedIn Data

LinkedIn ingestion via browser automation. No classification, no triage. Slow — requires browser. Cannot run in parallel with other browser-based syncs.

Output File

outputs/sync-data/linkedin.json — full replacement on every run.

Schema

{
  "version": 1,
  "last_synced": "2026-02-11T09:05:00-08:00",
  "conversation_count": 20,
  "conversations": [
    {
      "name": "Contact Name",
      "title": "Their LinkedIn title/headline",
      "isGroup": false,
      "lastActivity": "2026-02-11 08:30",
      "unreadCount": 0,
      "lastMessagePreview": "text",
      "isMessageRequest": false,
      "messages": [
        {
          "sender": "Contact Name",
          "timestamp": "2026-02-11 08:30",
          "content": "Message text here"
        }
      ]
    }
  ]
}

Steps

1. Get browser tab context via tabs_context_mcp.
2. Use an existing LinkedIn tab or create a new one. Navigate to https://www.linkedin.com/messaging/.
3. Wait for message list to load. If login required, ask Master to authenticate.
4. Extract the conversation list (first 20 conversations) from the sidebar.

For each conversation, capture list-level metadata:

• name: contact name (or group name)
• title: their LinkedIn headline/title if visible in the sidebar
• isGroup: true if group conversation indicator present
• lastActivity: timestamp of last message
• unreadCount: unread badge count (0 if none)
• lastMessagePreview: the preview text visible in the sidebar

5. Open EVERY conversation and extract messages. For each conversation in the list:

   a. Click on the conversation in the sidebar. b. Wait 1-2 seconds for messages to load. c. Verify the conversation opened by checking the header matches the expected contact name. d. Extract the last 5-10 visible messages. Each message: { sender, timestamp, content } e. If the contact's title wasn't visible in the sidebar, capture it from the conversation header.

6. Check "Message Requests" tab. After processing regular conversations: a. Look for "Message Requests" or "Other" tab/filter in the messaging sidebar. b. If found, click into it and scan for conversations with messages. c. Extract these the same way, with isMessageRequest: true. d. Include them in the conversations array.

7. Only include conversations currently visible. The output is a full replacement — do NOT carry forward conversations from previous syncs. Output reflects exactly what's in the message list now.

8. Build the output object with version, last_synced (current ISO timestamp with timezone offset), conversation_count, and conversations.

9. Write to outputs/sync-data/linkedin.json (full replacement).

10. Log: "Synced {N} LinkedIn conversations at {timestamp}"

Known Issues

• LinkedIn messaging UI changes frequently. If selectors break, use read_page or find tools to locate elements by their purpose rather than hardcoded selectors.
• LinkedIn may show "Premium" upsell overlays — dismiss them and continue.
• Message timestamps on LinkedIn are often relative ("2h ago", "3d") — capture as-is, don't try to convert.
• Some conversations may show "InMail" badge — still extract normally.
• LinkedIn virtualizes the conversation list — scroll down if fewer than 20 conversations are visible.

Notes

• This skill does NOT classify or triage. That's alfred-linkedin-message-triage.
• The file is always the latest snapshot. No historical accumulation.
• Requires exclusive browser access — cannot run in parallel with other browser-based skills.