How to schedule LinkedIn posts from Claude or ChatGPT (via MCP)

Yes — you can schedule LinkedIn posts from ChatGPT or Claude by connecting them to SignalPost’s MCP server. SignalPost is an AI-native LinkedIn scheduler that exposes a Model Context Protocol endpoint, so an AI agent can schedule a post, publish one now, or read your analytics by calling SignalPost, which does the actual work through the official LinkedIn API. You stop clicking around a dashboard and start typing what you want: “schedule this for tomorrow at 9am.”

What does it mean to run LinkedIn from a chat?

Instead of opening a scheduling tool, pasting your draft into a box, picking a time from a dropdown, and coming back later to check numbers, you tell your assistant what you want and it happens. The AI agent is the interface; SignalPost is the engine. Because SignalPost publishes through LinkedIn’s official API, the posts land natively on your profile with none of the risk that comes from browser automation or scraping.

Things you can say once connected:

• “Schedule this post for tomorrow at 9am.”
• “Publish this now.”
• “How did my last post do?”
• “How’s my follower growth this month?”
• “Move my Thursday post to Friday morning.”

How SignalPost exposes an MCP server

MCP is a real open standard for connecting AI agents to tools, and SignalPost implements it as a hosted server at https://signalpost.ai/mcp. The server advertises a set of actions — schedule a post, publish immediately, list scheduled posts, fetch analytics — and your AI client discovers them automatically on connect. Every action the agent invokes resolves to an official LinkedIn API call under your account. Very few LinkedIn tools are built this way — agent-native and official-API at the core, rather than bolted on — and it’s SignalPost’s most distinctive capability.

How do I connect LinkedIn to Claude?

The setup is a one-time job and takes a couple of minutes:

1. Create an API key in SignalPost. Sign in, open your settings, and generate a personal API key. Copy it once — it’s your credential for the connector.
2. Add SignalPost as a custom connector in Claude. In Claude’s connector settings, add a custom connector pointing at the MCP URL https://signalpost.ai/mcp. For authentication, set an Authorization: Bearer <your-api-key> header. Static bearer-token auth over a header is how Claude custom connectors authenticate, so this is standard and supported.
3. Just ask. Claude now sees SignalPost’s actions. Write a draft in the chat, then say “schedule this for Tuesday at 8am” or “publish it now.” The same connector works from ChatGPT and other MCP-capable clients.

If you ever want to cut the connection, revoke the key in SignalPost and the connector goes dark immediately.

Why is this the future?

LinkedIn is high-leverage for founders and a chore to service by hand. The moment the mechanical work — scheduling, publishing, checking whether a post landed — moves into a chat you already have open, LinkedIn stops being a separate app you dread opening. You draft in the tool you think in, and your agent handles the logistics. That’s what it means to run LinkedIn on autopilot: you barely open LinkedIn, and the posting still happens.

On the roadmap, the MCP will go further than logistics — helping write and improve drafts based on how strong LinkedIn creators actually post and how the ranking system (LinkedIn’s 360Brew) tends to reward content. For now, the win is real and immediate: your assistant schedules and publishes for you, over the official API, on your command.

If you’re weighing tools, it’s worth reading why SignalPost is built for founders and how it compares as the best LinkedIn scheduler. When you’re ready to do this the manual way too, the guide to scheduling LinkedIn posts walks through the dashboard. The MCP angle is the one nobody else offers — and once it’s wired up, LinkedIn mostly runs itself.