---
source: https://autoadify.com/mcp/claude
title: "Autoadify in Claude"
---

# Schedule social media posts from Claude

> Add https://mcp.autoadify.com/mcp to Claude and sign in to Autoadify. Claude can then write posts, generate images and videos, and schedule them to Instagram, Facebook, LinkedIn, X, TikTok, YouTube, Threads, Bluesky, Pinterest and Reddit. No API key. Included on Autoadify plans from Pro.

Draft a launch post with Claude, have it make the image, and schedule it to LinkedIn, Instagram and X without leaving the chat.

## How to connect Autoadify to Claude

1. In Claude, open Settings → Connectors and click Add custom connector.
2. Name it Autoadify and paste https://mcp.autoadify.com/mcp.
3. Click Connect, sign in to Autoadify and approve the permissions.
4. In a chat, switch Autoadify on from the + menu → Connectors.

**Where it works:** claude.ai, the Claude desktop and mobile apps, and Cowork. Free Claude accounts can add one custom connector.

- On Claude Team and Enterprise, an owner adds the connector under Organization settings → Connectors first; each member then clicks Connect.
- The same connector works in Claude Cowork, so a long-running task can plan and schedule a week of posts.

## What Claude can do once connected

- `whoami` (read) - The workspace, plan, permissions and remaining quota of the key.
- `list_channels` (read) - Connected channels with each platform's posting rules — media required, image limits, text length, required fields.
- `create_post` (publish) - Schedules a post to one or many channels, publishes it now, or saves it as a draft for you to review. Per-platform text, media, Reddit and TikTok options.
- `list_posts` (read) - Posts with their status on each channel, filtered by status, channel or date.
- `get_post` (read) - One post: status per channel, the live URL, and the error if a channel failed.
- `update_post` (publish) - Edits a scheduled post or draft: text, per-platform text or time.
- `delete_post` (publish) - Cancels a scheduled post.
- `list_media` (read) - Searches the media library by name, alt text and a description of what each file shows.
- `get_media` (read) - One file: status, URL, type, size and dimensions.
- `import_media` (publish) - Pulls an image, video or PDF from a public URL into the library.
- `create_media_upload` (publish) - A one-time URL for uploading a local file, for agents with a shell.
- `complete_media_upload` (publish) - Finishes an upload once the file has been sent.
- `delete_media` (publish) - Deletes a file; refused while a scheduled post still uses it.
- `list_ai_models` (generate) - The image and video models, their credit cost, which ones your plan includes, and your credit balance.
- `generate_image` (generate) - Makes an AI image and saves it to the media library, ready to attach to a post.
- `generate_video` (generate) - Starts an AI video; it lands in the media library in 1–10 minutes depending on the model.
- `get_generation` (read) - Whether an image or video is done, with the media id to attach. Failed ones are refunded.

You choose Read, Generate and Publish when you approve. Ask for drafts to review posts in Autoadify before they go out. Disconnect any time from the AI assistants page in Autoadify.

Other assistants: [ChatGPT](https://autoadify.com/mcp/chatgpt), [Grok](https://autoadify.com/mcp/grok), [Perplexity](https://autoadify.com/mcp/perplexity), [Manus](https://autoadify.com/mcp/manus), [Claude Code](https://autoadify.com/mcp/claude-code), [Codex](https://autoadify.com/mcp/codex), [Cursor](https://autoadify.com/mcp/cursor), [Gemini CLI](https://autoadify.com/mcp/gemini), [OpenClaw](https://autoadify.com/mcp/openclaw), [Hermes Agent](https://autoadify.com/mcp/hermes).

---

_Source: https://autoadify.com/mcp/claude_
