add_media
ChatGPTStore one file in the media library and append it to posts, in one call. Exactly one source: file, url, upload, or media_ids / media_urls (already in the library, from list_media). attach_to_post_ids lists every post it goes on; nothing already there is removed (attach_media sets the exact list). The result names the posts that got it (attached), those that did not and why (notAttached), and those that still cannot publish (publishBlockers). A file is never read into the call as base64. PNG, JPEG, GIF, WebP, AVIF images up to 10 MB; MP4, MOV, WebM video and PDF up to 1 GB by upload, 100 MB by url or file. Args: file: A file the user attached to a ChatGPT conversation. ChatGPT fills it in and Antwork downloads it; it is never synthesized. ChatGPT's mobile apps send no usable file, which the result says. url: Public HTTPS URL of the file (e.g. an AI-generated image already hosted). Antwork downloads it. upload: A file in a shell or code-execution sandbox: {} returns a one-time uploadUrl (10 minutes), and curl -sS -X PUT --data-binary @<path> '<uploadUrl>' stores and attaches it; that response is the result. mime_type and byte_size are optional. media_ids: Library media ids to add to the posts. media_urls: Library media URLs to add to the posts. attach_to_post_ids: Every post to add the media to. workspace_id: Workspace ID or name. Optional if the user has only one.
attach_media
ChatGPTSet the exact media list on a post: reorder, remove or clear. Replaces the post's media with media_urls in that order; an empty list removes all of it. Appending is add_media. PDFs render only on LinkedIn (document posts). Leaving a post without media its platform requires still saves, with a publishBlocker in the result. Args: post_id: The post ID to set media on. media_urls: Media URLs from list_media (or this workspace's storage). workspace_id: Workspace ID or name. Optional if user has only one workspace.
create_campaign
ChatGPTDraft one idea across several accounts in one call: one independent draft per account, grouped as a campaign. Nothing is sent. Each variant carries its own copy for its account; there is no shared-body parameter. media_urls goes on every draft. The result's posts has one row per variant with its postId and links, or the reason it was not created (the others are still created). schedule_post / publish_post with the campaignId sends the set. Args: variants: One per account: {"account_id": str, "text": str} plus optional hashtags, goal, user_tags, pinterest_board_id, platform_options, as in create_post. campaign_id: An existing campaign to add these to. Omitted = a new one. media_urls: Library media URLs attached to every post in the campaign. scheduled_for: ISO 8601. Proposes a time on every draft; nothing is scheduled. workspace_id: Workspace ID or name. Optional if user has only one workspace.
create_post
ChatGPTSave ONE draft for ONE social account. Nothing is sent: schedule_post or publish_post does that. Several versions for the same account are one call: text is version A, variants up to 3 more, each its own draft. One idea on several accounts is create_campaign. The result carries characterCount / characterLimit (overLimit when the text cannot publish), and, when they apply: mediaRequired (TikTok and YouTube need a video, Instagram and Pinterest an image or video), derivedTitle (YouTube takes the first 100 characters unless platform_options.title is set), settingsRequired (TikTok's audience is chosen by the user at settingsUrl) and pinterestBoard. Args: text: Post copy. @mentions in text auto-link on X, Threads, TikTok and Facebook; LinkedIn rich mentions and Instagram photo tags go in user_tags. account_id: Social-account id (from list_social_accounts), not a platform name. hashtags: Appended to the text as #tags, counted in its limit (YouTube: video tags). A list or comma/space-separated string. goal: e.g. "Educate", "Engage", "Promote", "Entertain". scheduled_for: ISO 8601. Proposes a time on the draft; only schedule_post schedules. campaign_id: Add this draft to an existing campaign. user_tags: [{username, platform, platformId (LinkedIn URN), type ("mention" | "photo_tag"), x, y (0-1, Instagram photo tags)}]. pinterest_board_id: Pinterest only, from boards in get_post_context. Omitted = the account's first board. platform_options: Only what the user asked for. Instagram: publishType ("timeline" | "story"), postToGrid. Facebook: publishType ("timeline" | "story"; a story is one photo or one video, its text is not shown). Pinterest: pinTitle (≤100), link (the pin's destination URL, https). YouTube: privacyStatus ("public" | "unlisted" | "private"), title (≤100), categoryId (e.g. "22", "28"), madeForKids. TikTok settings are set by the user at settingsUrl, not here. variants: Up to 3 alternative versions for the same account. workspace_id: Workspace ID or name. Optional if user has only one workspace.
create_workspace
ChatGPTCreate a new workspace owned by the authenticated user. Use this when the user wants a separate brand / client / project space — each workspace has its own connected social accounts, posts, and media library. Subsequent tool calls can target the new workspace by passing workspace_id (its returned ID or its name). Args: name: Workspace name (required, 1–60 characters).
delete_media
ChatGPTDelete a media item from the workspace media library. This removes the Firestore record. The file in storage may still exist but will no longer appear in the media library. Args: media_id: The media item ID to delete. workspace_id: Workspace ID or name. Optional if user has only one workspace.
delete_post
ChatGPTDelete a post. Drafts and scheduled posts are removed; a published post is hidden but keeps its metrics history unless preserve_analytics=false. Args: post_id: The post ID to delete. delete_from_platforms: True also deletes it from every platform it was published to; a list of platform names limits that to those. TikTok has no delete API. Omitted, the live posts stay. preserve_analytics: Published posts only; defaults to true. workspace_id: Workspace ID or name. Optional if user has only one workspace.
fetch_platform_posts
ChatGPTRecent posts read back from the platforms themselves (not only Antwork's), with metrics and permalinks. One entry per account in accounts, so two LinkedIn accounts come back side by side. LinkedIn personal profiles cannot be read this way (LinkedIn closed that API): they come back unsupported: true with a reason. Antwork-published posts on them still have metrics in get_performance and get_post. Args: platform: One platform, e.g. "linkedin". Omitted = every connected one. account_id: One connected account. max_posts: Maximum posts per account (1-25, default 10). workspace_id: Workspace ID or name. Optional if user has only one workspace.
get_connection_urls
ChatGPTGet the links to connect or re-auth social accounts. Returns an OAuth connect URL per platform, for the user to open. This is the tool for MANAGING connections — connecting a new account, reconnecting a stale one, or seeing the connect options. For reading what's already connected, list_social_accounts is the lighter call. Args: workspace_id: Workspace ID or name. Optional if user has only one workspace.
get_engagement_history
ChatGPTEngagement earned per day, as a long-format table: one row per (date, platform), plus an "all" row per date. Columns: date, platform, likes, comments, shares, impressions, reachBasis, reach, saves, clicks, bookmarks, linkClicks, profileVisits, follows, watchTimeSeconds, degraded, awaitingPermission. A null is a day not measured or a field the platform does not report (draw a gap), a 0 is a measured zero. reachBasis names what impressions counts; different bases do not add, so the "all" row's impressions is null on a mixed day and its reach is always null. degraded marks zeros that are artefacts. Also returns platformTotals, rangeTotals, startDate / endDate. A message saying there is no daily data yet is the answer for any days value. followers is a second table, one row per (date, account): followers, dailyChange, approximate, unavailableReason (null followers, never 0). With post_id, one post instead: cumulative and daily columns per platform for every counter above, spanDays (a daily value can cover several days) and unreadableFields; without days, its whole history. Args: days: Trailing window in days (1–365, default 30). Includes today. post_id: One post's history instead of the workspace's. workspace_id: Workspace ID or name. Optional if user has only one workspace.
get_more_tools
ChatGPTCheck for additional tools whenever your task might benefit from specialized capabilities - even if existing tools could work as a fallback.
get_performance
ChatGPTLifetime engagement per published post, as a table, with totals per platform. {schema, rows, rowCount, platformTotals}, one row per post, with totalEngagement (likes+comments+shares) for ranking. Also reach, saves, clicks (any click), bookmarks, linkClicks (to the post's URL), profileVisits, follows (gained), watchTimeSeconds and avgWatchTimeSeconds. impressions means a different quantity per platform (impressions, views, reach, a rolling 30-day window), named by each row's reachBasis, so rows with different bases do not add or compare; reach never adds either. A blank cell is a metric the platform does not report for that post, not a zero. degraded lists counters carried forward from a failed read. Args: limit: Max rows to return (default 30, max 50). days: Posts published in the last N days (default 30, max 365). platform: Restrict to one platform. workspace_id: Workspace ID or name. Optional if user has only one workspace.
get_post
ChatGPTOne post in full: text, account, status, schedule, media, links, and for a published post its lifetime metrics. metricCells holds each metric (likes, comments, shares, impressions, reach, saves, clicks, bookmarks, linkClicks, profileVisits, follows, watchTimeSeconds) with whether it was measured (a 0 the platform reported vs a field it does not report). The text is current even if the user edited it in the web composer (editUrl; mediaUrl and settingsUrl open its media picker and platform settings). Per-day history is get_engagement_history(post_id=...). Args: post_id: The post ID. workspace_id: Workspace ID or name. Optional if user has only one workspace.
get_post_context
ChatGPTThe drafting bundle for one account, or several: workspace, account, and the account's recent posts. - workspace: name, website, logoUrl, brandVersion, timezone, preferredTimes, tagOutboundLinks. - account: id, name, handle, token health, dailyPostCap (posts per rolling 24h on the owner's plan, scheduled ones included), for Pinterest its boards, for Instagram recentTags (handles it tagged). - recentPosts: up to 15 of the account's published posts with their metrics. They are the account's voice. Empty means no history. - performance: what performed, from Antwork's measured metrics over 90 days: topPosts (posts published through Antwork, best first by rankedBy, the metric that matters on that platform: saves on Instagram, link clicks on Pinterest, watch time on YouTube...) and followers (count, change). Lean on what the top posts share (topic, hook, format) when choosing what to write. Absent when nothing is measured yet. Args: platform: "linkedin", "x", "facebook", "instagram", "threads", "tiktok", "youtube", "pinterest", "bluesky", "discord", "mastodon" or "telegram". Implied by account_id. With neither, the error lists the connected platforms and accounts. account_ids: Several accounts at once (a campaign): an accounts list, each with its own recent posts (fewer per account). Overrides platform / account_id. account_id: One account. Needed when the platform has several: otherwise error: "account_ambiguous" lists them. workspace_id: Workspace ID or name. Optional if user has only one workspace.
list_media
ChatGPTThe workspace media library: images, videos and PDFs, with URLs to attach. libraryUrl is the library page in the web app. nextCursor, when not null, is passed back as cursor for the next page. Args: media_type: "image", "video", "document" (PDFs), or omitted for all. limit: Max items per page (default 30, max 100). workspace_id: Workspace ID or name. Optional if user has only one workspace. cursor: nextCursor from the previous page.
list_posts
ChatGPTList the workspace's posts, optionally filtered, or laid out by day. Sorted scheduled (soonest first), then draft, published, failed. get_post(post_id) returns one in full. group_by_day=true is the calendar view: calendar maps each date (YYYY-MM-DD) to that day's posts by time, calendarUrl is the calendar page, and the range defaults to the Monday of this week through 30 days ahead. Args: status: "draft", "scheduled", "published", or "failed". query: Keyword matched against post text, case-insensitive. platform: e.g. "linkedin", "x". campaign_id: Only the posts of one campaign (from create_campaign). date_from: Scheduled on or after this date (ISO 8601, e.g. "2026-03-01"). date_to: Scheduled on or before this date (ISO 8601). limit: Max posts to return (default 50, max 50). Ignored by group_by_day. group_by_day: Group the posts by date instead of listing them. workspace_id: Workspace ID or name. Optional if user has only one workspace.
list_social_accounts
ChatGPTThe workspace's connected social accounts: the ids account_id takes. One entry per active connection: id, platform, name, handle, token health and a reauthUrl that refreshes its OAuth token. Accounts waiting on payment are listed apart in pendingPaymentAccounts. Connecting a new account is get_connection_urls. followers is the latest daily count (count, asOf) and its net change since changeSince over the last 30 days; a null count comes with a state/reason (e.g. awaiting_permission), never a 0. Args: workspace_id: Workspace ID or name. Optional if user has only one workspace.
list_workspaces
ChatGPTThe user's workspaces, their Antwork account (plan, post quota), and the active workspace's connected accounts. isDefault marks the workspace every tool uses when workspace_id is omitted. isOwner: false is a workspace the user was invited to; its owner pays for its accounts. account.quota.posts.remaining is what the plan still allows this cycle.
manage_routine
ChatGPTAutopilot: routines your AI runs on a schedule (e.g. "every Monday at 9, write 2 posts about X and schedule them"). action: list | get | create | update | pause | resume | delete | run_now. create needs name, brief (what each run posts), account_ids and cadence; mode defaults to "draft" (drafts for the user's approval); update takes any of them. Every answer carries the routine's sentence (the routine in one line), setup and autopilotUrl. Who starts each run (trigger.kind): - own_scheduler (default): the AI's own scheduled task runs it. Put setup.prompt in a recurring task on the cadence (ChatGPT Tasks, Grok Automations, Claude Code /schedule…); setup.steps says where. - claude_routine: Antwork starts the user's Claude routine. Needs its fire_url and token, which only the user can create (claude.ai/code/routines → Add trigger → API → Generate token). That Claude routine needs the Antwork connector and setup.prompt as its prompt. - chatgpt_events: Antwork sends routine.due; subscribe to it in this chat with setup.prompt. - webhook: routine.due to the workspace's webhook endpoints (paid plans). Args: routine_id: For every action but list and create. cadence: {"kind": "weekly", "days": [1..7] (1 = Monday), "time": "09:00", "timezone": "Europe/Madrid"} or {"kind": "every_n_days", "every_days": 2, "time", "timezone"}. mode: "draft" (drafts for the user's approval), "schedule" or "publish" (as the user chose when setting it up). max_posts_per_run: Most posts one run may make (default 3). trigger: {"kind": ..., "fire_url"?, "token"?}. Omit to keep the current one. workspace_id: Workspace ID or name. Optional if user has only one workspace.
publish_post
ChatGPTPublish one post, several, or a whole campaign now, and wait for the platforms to answer (about half a minute). A post whose earlier publish failed is retried (requeued about a minute out, retried: true). Every post first passes the same checks: character limit, required media and settings, a connected account. One post answers with its live url; several come back per post in posts with publishedCount, and a post that cannot go is reported on its own while the rest still publish (at most 10 per call). A TikTok post answers tiktok_choices_needed until the user has answered each choice; never answer for them. Args: post_ids: The posts to publish. campaign_id: Every post of a campaign (from create_campaign) instead. post_id: One post; the same as post_ids=[post_id]. workspace_id: Workspace ID or name. Optional if user has only one workspace. tiktok_settings: A TikTok post's settings, as the user answered the tiktokChoices a first call returned: {privacyLevel, allowComment, allowDuet, allowStitch, commercialContent: {enabled, yourBrand, brandedContent}, userConsented: true}. Keyed by post id for several posts. Only the user's own answers for this post.
report_issue
ChatGPTReport what looks like an Antwork defect to the Antwork team, who read every report. For a result that contradicts what the call said it did, a user saying something did not happen, or a failure with no reason given. Not for a refusal that explains itself (limits, media_required, plan, reconnect). The calls made before it are attached by Antwork; describe what you saw. Args: what_happened: What went wrong, in plain words, with any error text seen. expected: What should have happened instead. tool_name: The Antwork tool that misbehaved, if one did. post_id: The post involved, if any. workspace_id: Workspace ID or name, if one was involved.
schedule_post
ChatGPTSchedule one post, several, or a whole campaign to publish automatically at scheduled_for. Every post first passes the same checks as publish_post. Several posts come back per post in posts with scheduledCount; one that cannot be scheduled is reported on its own and the rest still go (at most 10 per call). Different times per post are separate calls. Args: scheduled_for: ISO 8601 datetime (e.g. "2026-04-01T10:00:00Z"), in the future and at most 30 days ahead. post_ids: The posts to schedule. campaign_id: Every post of a campaign (from create_campaign) instead. post_id: One post; the same as post_ids=[post_id]. workspace_id: Workspace ID or name. Optional if user has only one workspace. tiktok_settings: As publish_post's.
set_default_workspace
ChatGPTSet (or clear) the workspace every tool uses when workspace_id is omitted. The default lives on the user's account, so it holds across conversations and clients; an explicit workspace_id on a call still overrides it. The result carries the workspace's connected accounts. Args: workspace_id: Workspace ID or name. Null or empty clears the default.
update_post
ChatGPTChange a draft or scheduled post; returns it as saved, with its links. A published post cannot be edited. account_id moves the post to another account (its platform follows). The same idea on another account with its own copy is a new draft. Args: post_id: The post ID to update. text: New post text. account_id: A different social account id to publish to. hashtags: Appended to the text as #tags, counted in its limit (YouTube: video tags). A list or comma/space-separated string. goal: Updated goal. status: "draft" or "scheduled". scheduled_for: ISO 8601 datetime for scheduling. Scheduling or moving the time needs the publish scope. campaign_id: Grouping id shared by related drafts. pinterest_board_id: Pinterest only, an id from boards in get_post_context. platform_options: Settings to change, merged into the post's own (keys as in create_post); null resets one to the platform default. Pinterest also takes newBoardName, which creates that board and sets it. TikTok settings are set by the user at settingsUrl. workspace_id: Workspace ID or name. Optional if user has only one workspace.
update_workspace
ChatGPTUpdate the workspace's identity (name, website, logo) or its settings. Only the fields passed change. The result's workspace is the workspace as saved, settings included. Args: name: The workspace / brand name (60 characters max). website: Brand website URL. logo_url: URL to the brand logo image. timezone: IANA timezone for scheduling, e.g. "Europe/Madrid". preferred_times: Preferred posting times as HH:MM, e.g. ["09:00", "18:00"]. tag_outbound_links: Append UTM parameters to links on the workspace's own website when a post publishes, so clicks stop reading as "direct" traffic. Needs website set; no other site's links are touched. workspace_id: Workspace ID or name. Optional if user has only one workspace.