Skip to main content
Schedule and publish from Claude, Cursor, Codex or any MCP client.
In short: PostMesh runs a remote MCP server at https://postmesh.app/api/mcp. Create an API key in Settings, API & MCP, add the server to your assistant with that key, then ask it to post. Assistants send text posts to Facebook Pages, Threads, Reddit, X, LinkedIn, Bluesky, Mastodon, Telegram and Discord.

​
How it works

Your assistant, PostMesh, your accounts

Your assistant, PostMesh, your accounts · 31 tools

1

Create an API key

Settings, API & MCP. The key is shown once; copy it then. See API keys.
2

Add PostMesh to your assistant

Use the setup for your client below.
3

Ask

For example: “Schedule this to LinkedIn and Threads for Friday at 10am.”

​
Setup by client

Claude Desktop and claude.ai: open Settings, Connectors, Add custom connector. Name it PostMesh and paste the connector URL shown in PostMesh’s Settings, API & MCP.
A key that may post sends posts right away, with no second confirmation, unless it asks first. Keep it private and revoke it if it leaks. New keys can read and write drafts but not post; see API keys.

​
What assistants can do

  • Publish or schedule text posts to Facebook Pages, Threads, Reddit, X, LinkedIn, Bluesky, Mastodon, Telegram and Discord
  • Read your connected accounts, scheduled posts and history
  • Cancel a scheduled post, retry a failed one and manage drafts
  • Start a draft from one of your templates
  • Check a post against each network’s limits before sending
To approve each post before it goes out, switch on Ask first for the assistant’s key. See Ask before posting.
Instagram, TikTok and YouTube need a photo or video. A key with the media scope can bring one in: create_upload_url gives a 10-minute link for one file of an exact size (up to 1 GB, less for photos), then finish_upload checks it; attach_media_from_url downloads a public link of up to 80 MB within 60 seconds. Both return a media_ref to pass in media on schedule_post or publish_post_now, or to save in a draft. generate_image makes a picture from a description and saves it to your media library, with a media_ref the same way; each image counts toward your plan’s monthly AI images.
Every tool and its parameters: MCP tool reference.