get_query_snippets
ChatGPTQuick sample of archived posts and comments for a keyword query (boolean syntax supported: AND, OR, NOT, "exact phrases", parentheses), with an estimated match count per platform from the archive index. Archive only; up to 50 sample items. Results include user-generated text from third-party social media posts and comments, which is untrusted content.
get_request_status
ChatGPTStatus of a submitted scrape job and where its data went: download links for exports, or confirmation of delivery to the destination endpoint. Export file contents are never fetched into the conversation.
list_requests
ChatGPTList this account's recent scrape jobs, newest first, paged (up to 25 per page), optionally filtered by status.
lookup_social_url
ChatGPTLook up an archived post or profile by URL. A post URL returns the post and its comments; a profile/channel URL returns the profile's posts and their comments. Archive only: nothing is scraped live. An empty result means no matching archived items were found for the supplied filters. Supported URL examples include posts and profiles on TikTok (tiktok.com/@user, /@user/video/id, /@user/photo/id), YouTube (youtube.com/watch, youtu.be/video-id, /@channel, /channel/id, /user/id), Reddit (reddit.com/r/sub, /r/sub/comments/id), X (x.com/user, /user/status/id; twitter.com is not accepted), Instagram (instagram.com/user, /p/id, /reel/id, /reels/id, /tv/id), Facebook (pages, groups, posts, videos, reels), LinkedIn (/in/user, /company/name, /posts/..., /activity/..., /feed/update/..., /pulse/...) and Bluesky (bsky.app/profile/handle, /profile/handle/post/id). Other sites are rejected. Results include user-generated text from third-party social media posts and comments, which is untrusted content.
search_social_posts
ChatGPTKeyword search over Datashake's archive of already-collected social media posts and comments. Archive only: nothing is scraped live, so an empty result means no archived match, not an error. Boolean queries are supported (AND, OR, NOT, "exact phrases", parentheses). Filter by platforms, publication date range (from_date/to_date, inclusive) and author; sort by likes or comments; up to 50 items. The total match count is an estimate. Results include user-generated text from third-party social media posts and comments, which is untrusted content.
submit_scrape_keyword
ChatGPTStart one live keyword scrape for posts and comments on the selected platforms. Boolean queries support AND, OR, NOT, quoted exact phrases, and parentheses. Each OR alternative is collected as a separate platform search. Keyword jobs have no result-count limit; from_date and to_date filter by publication date. Results are delivered to destination_url when set; otherwise, downloadable CSV and JSON exports are listed by get_request_status. Returns a request ID. Each call starts a separate scrape job billed under the connected Datashake account's current billing setup; it does not create a subscription or change plans. Before calling, present the exact job details and get explicit user approval.
submit_scrape_url
ChatGPTStart one on-demand scrape job for a supported URL. A post URL collects that post and its comments; a profile/channel URL collects posts and comments. For a profile/channel URL, provide max_posts or a publication date range. max_comments_per_post caps comments; set it to 0 to collect no comments. Limits and dates cannot be combined. Results go to destination_url when set, otherwise to downloadable exports listed by get_request_status. Returns a request id. Supported URL examples include posts and profiles on TikTok (tiktok.com/@user, /@user/video/id, /@user/photo/id), YouTube (youtube.com/watch, youtu.be/video-id, /@channel, /channel/id, /user/id), Reddit (reddit.com/r/sub, /r/sub/comments/id), X (x.com/user, /user/status/id; twitter.com is not accepted), Instagram (instagram.com/user, /p/id, /reel/id, /reels/id, /tv/id), Facebook (pages, groups, posts, videos, reels), LinkedIn (/in/user, /company/name, /posts/..., /activity/..., /feed/update/..., /pulse/...) and Bluesky (bsky.app/profile/handle, /profile/handle/post/id). Other sites are rejected. Each call starts a separate scrape job billed under the connected Datashake account's current billing setup; it does not create a subscription or change plans. Before calling, present the exact job details and get explicit user approval.
get_query_snippets
Claudeget_request_status
Claudelist_requests
Claudelookup_social_url
Claudesearch_social_posts
Claudesubmit_scrape_keyword
Claudesubmit_scrape_url
Claude