Tell my agent

๐Ÿ”– Stop sending links to yourself

  • ๐Ÿ“ฑYou save tweets about new tools in your bookmarks or your WhatsApp. You never open them again.
  • โœ๏ธInstead, reply @tell_my_agent under the post. Add a question if you want.
  • ๐Ÿค–It goes straight to your own AI agent, the one that already knows your project.
  • โœ…Your agent tells you if it's useful, how to use it and how long it takes. It can open a GitHub issue for you.
Sign in with X

โ‚ฌ10 a month, VAT included

Cancel in one click. Works with Claude Code routines, n8n, Zapier or any webhook.

One reply instead of one more bookmark

You stay on X. Your agent does the reading, with your codebase in front of it.

๐Ÿ‘คSomeone @someone

Just shipped an open-source job queue that runs on plain SQLite. No Redis, no broker. 2k jobs/s on a $5 VPS.

๐Ÿ™‚You

@tell_my_agent could this replace Celery in my app?

You see a post. You reply with the handle and, if you like, a question.

Tell my agent @tell_my_agent

โœ… Sent to your agent

The post, its links and your question reach your agent. You keep scrolling.

Replace Celery with the SQLite job queue

Verdict
Try it
Effort
About 3 hours
Why
Your 4 Celery tasks are simple and Redis only serves them.
First step
Move the daily email job and compare.

Your agent answers with your project in mind, here as a GitHub issue. You decide what it does.

Simple pricing

โ‚ฌ10 / month

  • ๐Ÿ”– 300 posts a month, up to 30 a day
  • ๐Ÿค– Claude Code routine or your own webhook
  • ๐Ÿ“ฌ One message per post, or a daily digest
  • ๐Ÿงพ VAT included, cancel in one click
Sign in with X

Questions

Do you read or analyse my posts?

No. We pass the post to your agent and nothing else. Your agent and its prompt decide what to do with it.

Which AI agent does it work with?

Any Claude Code routine (Pro, Max, Ultra, Team or Enterprise plan), or anything that accepts a webhook: n8n, Zapier, Make, your own server. The guide walks you through both.

Will my agent obey instructions hidden in a post?

Posts are written by strangers, so we hand them over clearly marked as untrusted text, and the routine prompt we give you tells your agent never to follow them.

What if I tag something by mistake?

It counts as one post of your monthly 300. Your agent will tell you it's not relevant.

Can I stop the replies on X?

Yes. Choose a reply after every tag, once a day, or never, in your dashboard.