Post to social media from ChatGPT

Which ChatGPT plans can schedule through a custom MCP server

OpenAI separates reading from writing. Its article on developer mode says full MCP support, “including modify/write actions”, is rolling out in beta to ChatGPT Business, Enterprise and Edu plans, and that it is available on ChatGPT web.[1]

The same article says Pro users can connect MCP servers “with read/fetch permissions in developer mode”, and that full MCP is “only available to Business and Enterprise/Edu users, currently”.[1]

Creating or scheduling a post changes data, so it is a write action. By those rules it needs a Business, Enterprise or Edu workspace today. OpenAI marks the feature as beta and says functionality and permissions may change.[1]

OpenAI also runs a plugin directory that is “available across ChatGPT plans”. Postnavo is not released, so it is not listed there, and this page describes the custom-server route only.[5]

What a workspace admin has to do first

On Business plans only admins and owners can enable developer mode and deploy an app. Each of them turns it on for their own account, and they cannot turn it on for individual members.[1]

On Enterprise and Edu, admins can use role-based access to give developer mode to chosen members, who then switch it on under Settings, Apps, Advanced Settings.[1]

A new app starts as a draft. Only admins and owners can publish it for the workspace. After that ChatGPT keeps a “frozen” snapshot of its tools: later changes by the developer are not applied until an admin reviews and publishes an update.[1]

For a team that means one approval when Postnavo launches, and another whenever its list of tools changes.

What ChatGPT asks before it posts

OpenAI lists four permission settings for a connected app: Always ask, Allow read actions, Allow low-risk actions and Allow all actions.[3]

Before an action runs, ChatGPT may show an approval card that describes the app and the proposed action. Always allow is not offered as a persistent permission to members of a managed workspace.[3]

Its examples of actions that may need extra review include sending an email, message, comment or invitation. Some requests may be denied instead of showing a prompt.[3]

Postnavo is launching with an approval setting of its own on top. When it is on, a post an agent creates waits in Postnavo for your OK, whatever you allowed in ChatGPT.

What ChatGPT does on a schedule, and where it stops

ChatGPT has scheduled tasks. The number of active tasks depends on the plan: 3 on Free and Go, 5 on Plus, 10 on Business and Edu, and 15 on Pro and Enterprise. Free accounts can repeat a task at most once a day, and eligible paid plans up to once an hour.[4]

A task does not push a post through by itself. OpenAI says an action that sends a message or changes external data may require approval, and that when it does, “the task pauses until you review it”.[4]

Two other features stay out of it. Agent mode “will not use custom apps”, and deep research can use them “for read/fetch actions only”.[1]

None of that decides when a post goes out. Postnavo is launching as the scheduler: ChatGPT hands over the text, the channels and the time in one conversation, and the post then sits in your Postnavo calendar until that time.

If your plan cannot write through ChatGPT

You can still draft in ChatGPT on any plan and schedule the result in Postnavo by hand. Postnavo is launching with a calendar and a composer, and with Navo, its built-in agent, which needs no ChatGPT subscription.

The connection itself is the Postnavo MCP server, the same one every other agent on this site uses.