X mention workflow

For tagged X moments, keep reply memes draft-only.

A mention comes in. The agent drafts reply text, renders a meme preview, and puts the whole thing in your approval queue.

Launch scope: draft-only by default. Auto-reply is founder-gated; customer brands stay human-approved.

See the approval loop Deploy the agent
Product composite of an X reply workflow with real rendered meme drafts and approval state
Product composite using the current approval model and real rendered meme PNGs captured from the production live quality battery.

The reply loop stays supervised.

Configured mentions

When configured, the bot watches for tagged opportunities in the AI/dev conversation and reconstructs enough context to draft a relevant reply.

Draft the reply

It prepares the reply text, meme preview, and share set together so the reviewer can judge the whole post, not just a caption.

Open, approve, post

Your queue shows the reply draft and media. A human opens the reply intent or posts from the approved draft.

1. X mention enters the draft queue.
2. Agent writes a reply and renders media.
3. Queue shows draft-only state.
4. Human approval before posting.

What this is not.

  • Not self-posting for customers. The product posture is assisted drafting and human approval.
  • Not a broad social listening suite. The launch wedge is AI/dev culture where the meme grammar and current examples are strongest.
  • Not a black box. The approval queue keeps reply context, draft text, media, and decision state visible.

Want reply drafts without losing the human checkpoint?

See the mechanism Back to AI/dev agent