Influence's hosted MCP server lets compatible AI assistants prepare social media posts, plan campaigns and read saved results from your Influence account.
Use your saved brand voice and audience to guide drafts. Tailor posts for each channel, choose media and set posting times in your time zone. Review and approve the final content and schedule in Influence before delivery. Review image posts in Influence before publishing.
Track delivery status and use Insights to review saved account posts and performance. The content calendar shows upcoming posts.
Connect X (Twitter), LinkedIn profiles and Pages, Instagram, Facebook Pages, Threads, YouTube, TikTok, Pinterest, Bluesky, Mastodon, Discord, Slack, Telegram, Twitch, Nostr, Tumblr, Pixelfed and WhatsApp Business. Influence also connects WordPress, DEV, Hashnode and listmonk.
MCP endpoint: https://influence.so/api/mcp Sign in with your own Influence account through OAuth or a scoped bearer key. Setup: https://github.com/influence-so/influence-mcp#connection
Influence is a paid subscription with a seven-day trial. A payment card is required for the trial. Paid plans include unlimited connected accounts for one business.
Worked publishing example, with review steps and screenshots: https://influence.so/publish-from-chatgpt
Scored across 65 tools
Most platform-specific list/search tools are distinguishable by provider, but the campaign lifecycle has several overlapping tools (campaignDraftSave, campaignReview, campaignReviewApprove, campaignReviewExecute, campaignScheduleBatchReview, campaignScheduleBatchExecute) that could be confused without careful reading. Media upload and generation tools are also stepwise and require understanding the pipeline.
Almost all tools use camelCase with a consistent resource/action pattern (accountList, campaignGet, mediaUploadCreate), which is predictable across the large surface. Minor deviations are bare generic verbs like fetch and search, plus a few noun-only names such as campaignReview.
With 65 tools, the surface is very heavy for an MCP server, even though the platform covers many social providers and complex publishing workflows. Many provider-specific list/search tools are individually justified, but the overall count creates selection burden and exceeds a well-scoped tool set.
The server covers account discovery, campaign planning/drafting/reviewing/scheduling/execution, media import/upload, insights, ads audits, brand context, and many platform destination lookups. There are minor gaps such as no obvious campaign deletion/cancellation or media deletion, but core publishing workflows are well represented.
Returns paginated owned accounts, stable identities and core-post eligibility. The platform field is a stable identifier; account eligibility determines availability. Provider labels: x=X; linkedin=LinkedIn profile; linkedin-page=LinkedIn Page; reddit=Reddit; instagram=Instagram via Facebook Business; instagram-standalone=Instagram Standalone; facebook=Facebook Page; threads=Threads; youtube=YouTube; gmb=Google My Business; tiktok=TikTok; pinterest=Pinterest; dribbble=Dribbble; discord=Discord; slack=Slack; kick=Kick; twitch=Twitch; mastodon=Mastodon; bluesky=Bluesky; lemmy=Lemmy; wrapcast=Farcaster; telegram=Telegram; nostr=Nostr; vk=VK; medium=Medium; devto=Dev.to; hashnode=Hashnode; wordpress=WordPress; listmonk=ListMonk; moltbook=Moltbook; whop=Whop; skool=Skool; mewe=MeWe; tumblr=Tumblr; pixelfed=Pixelfed; plurk=Plurk; ghost=Ghost; discourse=Discourse; misskey=Misskey; mattermost=Mattermost; matrix=Matrix; teams=Microsoft Teams; peertube=PeerTube; signal=Signal.
Inputs: limit, cursor. Returns: object
Returns one owned private ad review, actual credit cost, creative-quality scores, evidence and qualified comparison examples. Missing reports return null; unknown cost and unavailable comparisons stay explicit. Oversized reports include their signed-in review path.
Inputs: auditPublicId (required). Returns: object
Returns paginated saved private ad reviews with stable IDs and current job states. Saved reviews remain readable while new processing is unavailable.
Inputs: limit, cursor. Returns: object
Returns saved, confirmed company facts and current customer-authored brand guidelines. Missing context remains absent; this operation performs no website research.
Inputs: no named parameters. Returns: object
Creates one durable draft or patches selected destinations at its exact saved version. Supplied destination settings replace prior settings; omitted fields and other destinations stay intact. Drafts can omit required media until review. campaignScheduleBatchReview combines draft saving and review creation for 1–10 finished scheduled posts, including one post; this tool also supports draft-only work, edits and publish-now preparation. After preparation, confirm with only _confirmationCapability and _idempotencyKey; omit the draft fields to reuse the exact prepared draft.
Inputs: title, targets, timezone, campaignId, +3 more. Returns: object
Returns exact saved text, settings, ordered media and per-destination delivery status. Media retains its full verified facts in saved order; an exact single image/video also has its image/video reference.
Inputs: campaignId (required). Returns: object
Returns saved compositions with a continuation cursor.
Inputs: limit, cursor. Returns: object
Queues one AI text-planning turn using owned accounts, saved media and brand context, spending AI credits. Revisions require the plan identifier and current expectedVersion. Results are saved planning text and optional image ideas; this operation creates no image, video or audio and performs no publishing.
Inputs: brief (required), timezone (required), requestId (required), preferences (required), +9 more. Returns: object
Returns an owned AI plan, its exact version, conversation, saved image ideas, current planning job and available/reserved AI credits. Image job access depends on the current connection permissions and visible tools.
Inputs: planPublicId (required). Returns: object
Creates an exact pending campaign review, or saves explicit TikTok answers with reviewId and tiktokSettings to return a replacement pending review. Ask for all missing TikTok choices and applicable consent together; never infer answers. Saving choices does not approve or publish. Returns accounts, content, media, settings, time and review state.
Inputs: intent, reviewId, campaignId, scheduledFor, +4 more. Returns: object
Executes the exact reviewed campaign after an explicit click in the Influence review UI. Requires a private expiring capability for every exact review and all attached media to be inspected. Complete any pending TikTok choices before approval. Returns the completed review receipt or a rejected outcome with its reason; rejected means nothing was applied.
Returns: object
Executes an exact approved review once through the durable scheduler. Send approval: approved only after fresh explicit user approval of the complete post, media, settings and time. Chat approval is recorded as client-reported. TikTok Direct first requires explicit choices and consent saved through campaignReview or the review card.
Inputs: reviewId (required), approval, _idempotencyKey, _confirmationCapability. Returns: object
Returns an owned review, its complete content, current state and any committed operation result.
Inputs: reviewId (required). Returns: object
Schedules up to 10 exact approved schedule reviews atomically, with at most 50 deliveries. One explicit user approval can cover the complete displayed batch, including every destination, text, media, settings and time. Save any required TikTok choices and consent first. Invalid or stale items roll back new scheduling. Replays retain the committed job identifiers.
Inputs: items (required), _idempotencyKey, _confirmationCapability. Returns: object
Saves 1–10 finished posts and pending schedule reviews together, including one post, with at most 50 social destinations. Returns compact review receipts. Invalid items roll back the batch.
Inputs: items (required), _idempotencyKey, _confirmationCapability. Returns: object
Returns the current connection's permissions, renewal mode, connection and access-token expiry, registered callback class and last successful account read. A native callback does not identify a vendor.
Inputs: no named parameters. Returns: object
List organizations found in the connected account's article history. Results do not confirm current membership or publishing permission. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List available article tags. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List available Discord publishing channels. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List forum categories where the connected user can create topics. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List forum topic tags available to the connected user. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
Find a visible forum topic to reply to. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: q (required), connectionId (required), cursor. Returns: object
List available publishing teams. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
Read a workspace-owned brand, saved campaign, AI plan or cached Insights document by its search identifier, including its source URL and provenance. Oversized documents fail rather than silently losing content; no provider refresh or publishing.
Inputs: id (required). Returns: id, url, text, title, +1 more
List authors available for this publication. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List active newsletters. Selecting one explicitly sends email when the article publishes. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List article tags for this publication. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List membership tiers available for article visibility. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List available article publications. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
Read recent native image batches for an owned plan: queued/running/final outcomes, owned asset IDs, credit reservations and authoritative settled costs. Unknown cost stays null and needs_review is not success.
Inputs: planPublicId (required). Returns: object
Prepare a native expiring image quote for one saved AI image idea at the exact plan version and owned reference IDs. Show its estimated/maximum credit reservation before requesting imageGenerationStart. This does not call an image model or charge credits.
Inputs: ideaId (required), messageId (required), requestId (required), planPublicId (required), +5 more. Returns: object
Confirm the exact unexpired image quote token, reserve its maximum AI credits and queue native image jobs. Use imageGenerationList for outcomes. Generated image IDs do not approve publication. Show the actual images through mediaGet, then obtain fresh explicit approval of the complete publishing review before sending.
Inputs: requestId (required), quotePublicId (required), confirmationToken (required), _idempotencyKey, +1 more. Returns: object
Returns owned Insights accounts, available channel totals, dated reporting periods, refresh capability, selected read channel and freshness. The reason and nextAction explain missing or stale data. reconnect_social_account means reconnect the social account in Influence, not this AI connector. channel.reason distinguishes unsupported account totals from available post metrics. Missing metrics remain unknown rather than zero.
Inputs: limit, cursor. Returns: object
Returns paginated cached posts, including posts published outside Influence, linked Influence campaigns when available, metric breakdowns, reporting windows, captions and measured/fetched timestamps. Undated readings are included by default and excluded when includeUndated is false. The account reason, nextAction and lastCheck explain empty or stale cached results; an empty page alone does not prove there are no posts. Missing metrics remain unknown.
Inputs: includeUndated (required), connectionPublicId (required), limit, cursor. Returns: object
Queues a bounded provider feed refresh that updates cached post and available account Insights. A queued result is not a completed refresh. Slack, Mattermost and Microsoft Teams read the channel already selected in Influence; this operation performs no publishing or channel selection.
Inputs: connectionPublicId (required), _idempotencyKey, _confirmationCapability. Returns: object
Returns native platform publishing rules, image and video limits, maximum text length and required settings.
Inputs: platform (required), isPremium, connectionId. Returns: object
Find a destination community matching nonempty search text. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: q (required), connectionId (required), cursor. Returns: object
List available subscriber lists. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List available campaign templates. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List joined channels available to this publishing account. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
Returns one owned ready image or MP4. Compatible MCP clients receive JPEG/PNG pixels up to 6 MiB. All formats include a five-minute previewUrl; request mediaGet again if it expires. Prefer native media or review cards. Signed URLs must remain unchanged: if the host rewrites links, use reviewPath when returned, on this connection's trusted Influence origin. That path opens the signed-in media library, not a specific asset. Larger JPEG/PNG, GIF, WebP and videos return metadata without downloading their bytes; images support up to 20 MiB and compatible MP4 videos up to 5 GB. Reading media grants no publication approval.
Inputs: assetId (required). Returns: object
Imports an authorized image or video into the workspace library from public HTTPS url with optional name, or a native host file reference. JPEG/PNG/WebP images support up to 20 MiB; MP4, MOV, WebM and M4V videos in standard dynamic range (SDR) support up to 250 MiB. Local binary image descriptors support up to 3 MiB. mediaUploadCreate supports larger ready SDR H.264/AAC MP4 files up to 5 GB, subject to destination limits. An optional stable clientKey identifies video retries; otherwise the exact public URL or native file_id identifies them. Returns an owned ready image or a pollable video upload status, with assetId only when ready.
Inputs: url, file, name, upload, +3 more. Returns: object
Returns ready owned source, edited and converted media with an optional all, image or video filter. Omitting the filter returns the established JPEG/PNG image list. Upload tools add files to the same library. Search or the bounded scan can return fewer than limit items; continue with cursor until done is true.
Inputs: kind, limit, cursor, search. Returns: object
Cancels an unfinished owned upload and uses the existing storage cleanup. Completed uploads are preserved. No new confirmation is needed. Repeating cancellation does not create another operation.
Inputs: uploadId (required). Returns: object
Verifies the uploaded parts and starts the existing media preparation. Returns uploading, processing, ready or a terminal failure. Only a ready result with assetId is attachable to a draft. Replaying the same part manifest recovers completion.
Inputs: parts (required), uploadId (required). Returns: object
Authorizes an exact file once using ordered part checksums. For batches, prepare and confirm each file before preparing the next; preparations can expire. Choose 64 MiB parts for new videos; 8 MiB is the default. Accepts supported images up to 20 MiB and MP4, MOV, WebM or M4V videos in standard dynamic range (SDR). MOV/WebM/M4V and video conversion support up to 250 MiB. Ready H.264/AAC MP4 supports up to 5 GB; destination limits still apply (YouTube 5 GB, other platforms up to 1 GB or their tighter limit). Returns the first batch of signed PUT URLs. Transfer parts directly to storage with Content-MD5, request nextPart only when needed, then finish without another confirmation. Reuse clientKey, partSizeBytes and identical checksums to resume. Uploads last 24 hours; URLs can be renewed. No file bytes in tool arguments.
Inputs: name (required), clientKey (required), sizeBytes (required), contentType (required), +4 more. Returns: object
Returns an owned upload, verified uploaded parts for recovery, processing status and the final ready assetId. It never returns write URLs or authorizes an upload. A converted asset can have a different ID from uploadId.
Inputs: uploadId (required). Returns: object
Returns up to 32 short-lived PUT URLs for the exact part checksums authorized when this owned upload was created. Use nextPart for later pages or startPart to renew expired URLs. No new confirmation is needed; the file manifest cannot change. Send Content-MD5, never Influence bearer credentials, to storage.
Inputs: uploadId (required), limit, startPart. Returns: object
List available publishing groups. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List available publishing boards. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
Returns the authorized account email and display name, workspace display name in nickname, and opaque workspace profile ID. The profile ID stays stable across refresh, reconnect and display changes.
Inputs: no named parameters. Returns: id, name, email, nickname
Read the selected subreddit's publishing requirements. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: subreddit (required), connectionId (required), cursor. Returns: object
Find a destination subreddit. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: word (required), connectionId (required), cursor. Returns: object
Returns compact paginated calendar entries in an inclusive UTC date window, ordered by scheduled time. Entries contain campaign identity, status and timing without content or media. Reading available times does not reserve them.
Inputs: endAt (required), startAt (required), limit, cursor. Returns: object
Search confirmed brand context, the 20 most recent saved campaigns and AI plans, and available cached Insights in your Influence workspace. Read-only; this is a bounded snapshot, not an exhaustive archive or web search.
Inputs: query (required). Returns: results
List joined Slack publishing channels. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List channels in the connected team available to this publishing account. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
Read fresh creator identity, allowed privacy levels, interaction restrictions and maximum video duration. Use the returned postingConsent wording and links in one grouped question for missing Direct Post choices and consent. Do not preselect privacy or infer consent. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List the authorized profile and community destinations. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List available publishing companies. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List the selected company's publishing experiences. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: companyId (required), connectionId (required), cursor. Returns: object
List available article categories. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List available article types. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
List available article tags. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: connectionId (required), cursor. Returns: object
Find a destination channel using at least two characters. Requires an owned active account and current provider availability. Follow nextCursor until done, keeping the same connection and search fields.
Inputs: q (required), connectionId (required), cursor. Returns: object
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Schedule and publish posts to TikTok, Instagram, YouTube, LinkedIn and more from any AI agent. Connect your social accounts once, then let Claude, ChatGPT or any MCP client create, schedule, edit and review posts across every platform. Upload images and videos to your media library and attach them to posts. Signs in with OAuth; HubPost never sees your passwords. Also available in the [Claude directory](https://claude.ai/directory/hubpost) and as a [ChatGPT app](https://chatgpt.com/plugins/plugin_asdk_app_6ab141477a708191a7c5f68714ee7c39). Setup guide: [hubpost.co/mcp](https://hubpost.co/mcp?utm_source=glama&utm_medium=mcp-directory).
Connect any AI agent to 11+ social platforms: schedule, publish & track posts via hosted MCP.
SocialFaktory (https://www.socialfaktory.com) is a social media MCP server that lets your AI agent run a brand's social content with you. It reads the brands, channels and media you already have, writes posts in the brand's own voice, prices and generates short video, takes a file you upload, composes one post per channel, schedules or sends it on TikTok, Instagram, YouTube, X, LinkedIn, Facebook and Pinterest, and reads the metrics back. It is a hosted remote server at https://www.socialfaktory.com/mcp (Streamable HTTP) with OAuth 2.1 sign-in and 20 tools. Install steps cover Claude Code, Claude Desktop, Cursor, VS Code, Codex CLI, Gemini CLI and Windsurf. You stay in control. Generating spends the credits in your wallet, and the agent is told to quote the price and ask you first. Posts are drafts until you send them, and nothing reaches a channel without the publish permission you grant on the consent screen, where you also pin the connection to one brand, cap monthly spend and choose when it expires. Generating and publishing need an active SocialFaktory plan. Not available through an agent yet: cloning a video from a link, and generating still images or carousels. Connecting a social channel is a browser sign-in and stays in the app. Docs: https://www.socialfaktory.com/docs/mcp
Run your social media from Claude, ChatGPT or any MCP client. Feedlio's MCP server lets your assistant list workspaces and connected accounts, add media, and create, edit, reschedule and cancel posts across 15 platforms, including Facebook, Instagram, LinkedIn, YouTube, TikTok, Pinterest, Threads, Bluesky, Mastodon, Dev.to and Google Business Profile. Sign in with OAuth on Feedlio's own page. MCP is included on every Feedlio plan, the Free plan too. Support: support@feedlio.io. Also in Claude's Connectors Directory: https://claude.ai/directory/connectors/feedlio
Add one secure layer between your agents and this server.
