Announced October 7, 2026, OmniSocials exposed its social scheduling, publishing, analytics, and inbox stack to AI assistants — 48 tools, reachable from Claude, ChatGPT, Cursor, and any MCP client, included in its single $10/workspace plan.
2026-10-08 · by Moe Ameen
On October 7, 2026, OmniSocials — a Netherlands-based social media management platform founded by Robert Ligthart — announced a hosted Model Context Protocol (MCP) server that lets an AI assistant plan, schedule, publish, and analyze social content in plain language. The pitch, per the release, is removing the copy-paste step: assistants are already used to draft posts, but people then paste the text into another app to actually ship it. Connect OmniSocials once and the assistant can "publish the post, check how it performed, and reply to comments, all from the same conversation."
The server exposes 48 tools grouped into workspaces and accounts, posts, approvals, media, hashtag sets, discovery, analytics, a social inbox, and webhooks. An assistant can create posts, stories, and reels as drafts, scheduled items, or immediate publishes; run approval workflows; upload media (up to 100 MB per file, or 1 GB via URL, with PDFs turned into carousels); read 7-, 30-, or 90-day analytics and suggested best posting times; and work a social inbox — reading and replying to DMs, comments, and mentions. It covers 11 platforms: Instagram, Facebook, LinkedIn (profile and page), YouTube, TikTok, X, Pinterest, Threads, Bluesky, Mastodon, and Google Business, with Reddit and Snapchat listed as coming soon.
Setup works with any MCP-compatible client — OmniSocials names Claude (web and desktop), Claude Code, ChatGPT, Cursor, Windsurf, and others — via a hosted remote server at mcp.omnisocials.com over streamable HTTP, or a local server through the npm package `@omnisocials/mcp-server`. Remote clients authenticate with OAuth; others use an API key, with inbox and analytics access behind opt-in scopes. The MCP server is included at no extra cost in OmniSocials' single plan, which runs $10 per workspace per month billed annually (a bit more monthly) with a 14-day trial. The same operations are also available through a REST API, SDKs, and a CLI.
A note on timing: OmniSocials' MCP package predates this announcement by several months, so October 7 is best read as a formal launch and expansion rather than the first release. Treat the exact tool count, platform list, and pricing as a launch-window snapshot and confirm against OmniSocials' own docs before relying on any single figure.
It is worth being precise about what OmniSocials does and does not change. It is a publishing, scheduling, and inbox layer: the MCP server lets an assistant draft a caption and then ship it, read analytics, and reply to comments across 11 accounts. What neither OmniSocials nor the assistant behind it produces is the actual content a feed runs on — the captioned short-form video, the brand-exact carousel, the face-locked avatar clip, the quote graphic, the blog, the newsletter. A chat assistant writing a caption is one asset; a content operation is many formats, on brand, every week. [Kompozy](/) is the engine for that half: from a single source it generates [Persona Shorts](/glossary/persona-shorts) and avatar video, [Carousel Posts](/glossary/hyperframes) and images, Quote Graphics, Blogs, and Newsletters, every output governed by a written [Persona Brief](/glossary/persona-brief) so a dozen pieces still read as one brand.
The governance point deserves its own line. The flip side of an assistant that can publish and reply to 11 accounts from a chat window is that an off-key prompt can ship something off-brand in your name. Kompozy is built the opposite way — generation runs behind a per-post [review gate](/glossary/autopilot), and [Autopilot](/glossary/autopilot) fans the approved set across the eight social platforms plus blog and email on a schedule, so nothing goes out until it passes. If this week's "my assistant can run my socials now" story is one you want to react to, that is also a content opportunity: drop the take into Kompozy and it becomes a carousel on what agent-native publishing means, a captioned Persona Short explaining MCP in plain English, and a blog for the long-tail search — made, not just moved. For the broader landscape, see our coverage of [BulkPublish opening its stack to agents](/news/bulkpublish-opens-publishing-api-to-ai-agents) and the guide to [social media MCP servers](/guides/social-media-mcp).
On October 7, 2026, OmniSocials launched a hosted Model Context Protocol (MCP) server that lets an AI assistant run social media in plain language — planning, scheduling, and publishing posts, reading analytics and best-time suggestions, and managing a social inbox of DMs, comments, and mentions across 11 platforms. It exposes 48 tools, works with any MCP client (Claude, ChatGPT, Cursor, and others), and is included in the company's single plan at no extra cost.
OmniSocials lists 11 platforms: Instagram, Facebook, LinkedIn (profile and page), YouTube, TikTok, X, Pinterest, Threads, Bluesky, Mastodon, and Google Business, with Reddit and Snapchat marked as coming soon. Because it is an actively evolving product, confirm the current list in OmniSocials' own documentation before relying on it.
Not itself. OmniSocials schedules, publishes, and analyzes content and manages the social inbox; the drafting happens in the connected AI assistant, which writes the caption. Neither produces the finished short-form video, carousel, avatar clip, or on-brand image a feed runs on — that is a separate job, handled by a content engine such as Kompozy, which generates those formats and publishes them across platforms.
The MCP server is included at no extra cost in OmniSocials' single plan, which is $10 per workspace per month billed annually (slightly more billed monthly), with a 14-day free trial and no permanent free plan. Confirm current pricing on OmniSocials' site, since it can change.