Skip to content

[slack] fleet posting volume trips the workspace app-message cap: trim mirrors or upgrade plan #81

Description

@DJRHails

The solo preview Slack workspace the gantry fleet posts to is on the free plan, and Slack's unpublished anti-spam cap on app-posted messages tripped at ~09:20-10:45 UTC on 2026-07-28. Every write API returns message_limit_exceeded workspace-wide while reads keep working. Re-probes at 16:10 and 20:41 UTC (~10-11h after the trip) still fail — at current volume this is chronic, not a transient throttle.

Volume evidence (14 days to 2026-07-28, channel roots only — thread replies uncounted, so real totals are higher):

  • 3,852 bot messages ≈ 275/day; the trip day was pacing ~2× that.
  • Top sources are the fleet's own event mirrors, not the loop reports: #gantry-runs 1,755, #portcullis 1,031 (~72% combined), #loop-* ≈ 800.

Recommendation (pick one):

  1. Trim the #gantry-runs / #portcullis mirrors — batch them or mirror failures only. Cheapest, and removes ~72% of volume.
  2. Upgrade the workspace to a paid plan.

Until then, workers fall back to PR comments / run results for delivery — the slack skill now fails fast with exactly that guidance on this error (#80, merged). Diagnosis details: triage run, original report from primal-polite-cedar.

via gantry

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions