LinkedIn Get Direct Messages Actor Apify actor that retrieves all direct messages from your LinkedIn messaging inbox with comprehensive conversation details, participant information, and message metadata. ## Features - Retrieves all conversations from LinkedIn messaging inbox - Extracts detailed information for each conversation: - Thread URL - Participant LinkedIn profile URL - Last message content - Sender information (firstname, lastname, occupation, company) - Message direction (from me or from contact) - Read/unread status - Message timestamp - Uses LinkedIn session cookies for authentication - Follows the same authentication pattern as other Apify LinkedIn actors - Multi-selector strategy for robust element detection - Automatic scrolling to load all conversations ## Input The actor requires the following input parameters: ### Required - liAtCookie (string, required): Your LinkedIn li_at cookie value for authentication - Example: "AQEDATEtZ..." - How to get it: Open LinkedIn in your browser, open Developer Tools (F12), go to Application/Storage → Cookies → https://www.linkedin.com, and copy the value of the li_at cookie ### Optional - jsessionid (string, optional): Your LinkedIn JSESSIONID cookie value for additional authentication - Example: "ajax:1234567890123456789" - Found in the same location as li_at cookie - userAgent (string, optional): Custom user agent string - Default: "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36" - Use a recent Chrome/Edge user agent for best results ## Input Example json { "liAtCookie": "AQEDATEtZ...", "jsessionid": "ajax:1234567890123456789", "userAgent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36" } ## Output The actor outputs a single dataset item containing: - success (boolean): Whether the operation was successful - message (string): Status message - conversationCount (number): Total number of conversations retrieved - conversations (array): Array of conversation objects - timestamp (string): ISO timestamp of when the data was retrieved ### Conversation Object Structure Each conversation in the conversations array contains: - threadUrl (string): Direct URL to the conversation thread - linkedInUrl (string): LinkedIn profile URL of the contact - message (string): Content of the last message in the thread - lastMessageFromUrl (string): Profile URL of the last message sender - firstname (string): First name of the contact - lastname (string): Last name of the contact - occupation (string): Current job title/occupation of the contact - companyName (string): Company name where the contact works - isLastMessageFromMe (boolean): Whether the last message was sent by you - readStatus (boolean): Whether the conversation has been read (true = read, false = unread) - lastMessageDate (string): Human-readable date of the last message (e.g., "2h", "Yesterday", "Sep 01") - timestamp (string): ISO timestamp of the last message (if available) ## Output Example json { "success": true, "message": "Successfully retrieved 25 conversations", "conversationCount": 25, "conversations": [ { "threadUrl": "https://www.linkedin.com/messaging/thread/2-MGFjMTExN2YtOWI3NS00NmJlLWEyNjMtNjhlZDFjZWZkNDVkXzEwMA==/", "linkedInUrl": "https://www.linkedin.com/in/ACoAABJnny0ButYTqma7Tmq-SqQDoFJVPpqBIr8", "message": "Salut Pierre, j'ai vu que tu as jeté un œil à Hootsuite...", "lastMessageFromUrl": "https://www.linkedin.com/in/ACoAABJnny0ButYTqma7Tmq-SqQDoFJVPpqBIr8", "firstname": "Charlotte", "lastname": "Clemens", "occupation": "VP of Sales Enterprise Europe - Hootsuite & Talkwalker", "companyName": "Hootsuite", "isLastMessageFromMe": false, "readStatus": true, "lastMessageDate": "Sep 01", "timestamp": "2024-09-01T10:30:00.000Z" } ], "timestamp": "2025-01-15T14:23:45.123Z" } ## How It Works 1. Authentication: Sets LinkedIn session cookies (li_at and optionally JSESSIONID) 2. Navigation: Opens LinkedIn messaging page (https://www.linkedin.com/messaging/) 3. Conversation Loading: Automatically scrolls to load all available conversations 4. Data Extraction: For each conversation: - Extracts thread URL, message content, timestamp, and read status from the messaging list - Navigates to the contact's profile page - Extracts detailed profile information (name, occupation, company) 5. Output: Returns all conversation data in a structured format ## Performance - Execution Time: ~2-5 seconds per conversation (due to profile information extraction) - Scrolling: Automatically scrolls up to 5 times to load all conversations - Anti-Detection: Includes user-agent spoofing and webdriver masking ## Error Handling The actor will throw an error if: - The liAtCookie parameter is missing - LinkedIn returns a non-200 status code - The messaging page fails to load - Authentication fails (invalid cookies) ## Notes - Cookie Expiration: LinkedIn cookies expire periodically. If the actor fails, try getting fresh cookies from your browser. - Rate Limiting: LinkedIn may rate-limit requests if you run the actor too frequently. Use reasonable delays between runs. - Profile Privacy: Some profile information may not be available if the contact has restricted their profile visibility. - Headless Mode: The actor runs in headless mode by default for better performance. ## Security - Never share your li_at cookie publicly as it grants full access to your LinkedIn account - Use Apify's secret input feature to store cookies securely - Cookies are only used for authentication and are not stored or logged ## Use Cases - Export LinkedIn conversations for backup - Analyze messaging patterns and response rates - Integrate LinkedIn messaging data with CRM systems - Monitor unread messages and follow-ups - Generate reports on networking activities ## Related Actors - linkedin-send-dm: Send direct messages to LinkedIn profiles - linkedin-like-post: Automatically like LinkedIn posts - linkedin-comment-post: Automatically comment on LinkedIn posts - linkedin-quote-post: Repost LinkedIn content with your thoughts ## Support For issues or questions, please contact the development team. ## Version History - 1.0.0 (2025-01-15): Initial release - Basic conversation retrieval - Profile information extraction - Read status detection - Multi-selector strategies for robust element detection