From ChatGPT draft to Nostr queue
The writing already happens in ChatGPT. With the public API or the CLI, the same conversation can create scheduled Nostr posts instead of text to paste.
NostrAutomate Nostr with ChatGPT
Draft Nostr posts in ChatGPT, then let postqueen schedule and publish them. One conversation in, a full Nostr queue out.
Quick start
Point a custom GPT action at the public API, or let agent modes that run commands use the CLI for Nostr.
Custom GPT or CLI
# Custom GPT action: call the public APIPOST https://api.postqueen.ai/public/v1/postsAuthorization: your_api_key # Or in agent modes that run commands:npm install -g postqueenpostqueen posts:create --json nostr-week.jsonWhat ChatGPT can run on Nostr
Windows that fit Nostr
Tell ChatGPT when your followers are most active and it schedules posts into that window on Nostr, day after day, without reminders.
Media attached, sized right
Images upload in the same postqueen call, so Nostr posts arrive with their media in place instead of a bare link.
Drafts until you are ready
Ask ChatGPT for drafts only. They sit on the postqueen calendar unscheduled, so you can edit them and pick the time yourself, and nothing reaches your followers on Nostr until you do.
Three steps to a full queue
Discover Nostr
ChatGPT calls postqueen to list your connected keys and relays and sees what Nostr supports, so drafts start from real capabilities, not guesses.
Draft with ChatGPT
Give it the brief. The agent writes Nostr-shaped posts and proposes publish times that match when your followers are around.
Check and publish
Look over the queue on the postqueen calendar and change anything you want. postqueen publishes each post at its time, and ChatGPT can pull the results after.
What Nostr allows
ChatGPT reads these limits from postqueen before it composes, so a draft that would be rejected never reaches the queue.
- Identity
- A keypair (npub and nsec), with no email, signup or account approval
- Note length
- The protocol sets no character limit on text notes, though relays can cap event size
- Media
- Images and video are linked by URL and rendered inline by most clients
- Relays
- A note only exists on the relays you publish it to, so most people use several
- Threads
- Replies reference the parent event, so clients render them as native threads
- Portability
- The same key works in Damus, Amethyst, Primal and every other client
NostrEverything postqueen does on Nostr
Scheduling, AI drafts, analytics and team roles, in one guide.
F.A.Q.
Frequently asked questions
How does ChatGPT connect Nostr to postqueen?
Through the public API from a custom GPT action, or through the CLI in agent modes that execute commands. Both use the same API key and cover Nostr posts end to end.
Can ChatGPT schedule Nostr posts at exact times?
Yes. Give it exact slots or a rule like mornings at nine and it sets publish times on the posts it creates, so your followers see them at the right moment.
Will ChatGPT publish to Nostr on its own?
Only if you ask it to. Tell it to save drafts and nothing is scheduled: the posts wait on the calendar until you set a time. Tell it to schedule, and they go out at the times you can see, which is the setup most people move to once they trust the output.
Can one brief cover Nostr and other channels too?
Yes. postqueen connects 30 channels, so ChatGPT can turn the same brief into Nostr posts plus adapted versions for every other channel you run.
Does ChatGPT access cost extra on postqueen?
No. The agent uses your regular API key and plan limits: unlimited posts on every plan, and channel counts from 5 on Creator up to uncapped on Ultimate.
Can AI write my Nostr notes?
Yes. The built-in assistant drafts and rewrites notes in your voice from a short prompt, and nothing goes out until you schedule it. Agents like OpenClaw or Claude Code can prepare and queue a whole batch through the CLI or the built-in MCP server.
What analytics does postqueen show for Nostr?
postqueen records every note it published and when, so your own history is complete. Engagement is a different matter: reactions, reposts and zaps are themselves events scattered across relays, so any count depends on which relays you read from and no number is authoritative. That is a property of the protocol rather than a gap in the tool, and your client will show you as much as any dashboard honestly could.
Ready to get started?
Queue this week's posts in minutes, or hand the whole thing to your favorite AI. postqueen keeps publishing either way.
Seven days free. Cancel in one click from settings.