All your agents are going async

· ai programming · Source ↗

Article

TL;DR

Agents outliving their callers is real — but the problem is interaction models, not HTTP.

Key Takeaways

  • Real problem: agents finish at 3am, callers want a Slack message, not a rejoined session.
  • Disclosure buried: author is building Ably — HN flagged this as a sales pitch.
  • Simplest real solution per community: $5/month VPS, SSH, tmux — no new protocol needed.

Discussion

Top comments:

  • [bozdemir]: Transport isn’t the problem; interaction model is — and it’s a buried Ably pitch

    The transport isn’t really the problem though. The interaction model is. When my cron agent finishes a task at 3am I don’t want a live session I can rejoin, I want it to drop a message in Slack or email me and shut up.

  • [TacticalCoder]: Entire async agent problem solved by SSH and tmux for five dollars monthly

    $5 per month dedicated server, SSH, tmux.

  • [edg5000]: Nothing wrong with HTTP; industry obsession with conversation streams is the real issue
  • [pdp]: Concurrent agents injecting messages into each other’s context doesn’t work well

Discuss on HN