You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Periodical monitoring status pushed to external systems? #308
Can we get a way to send x min status updates (or when something has changed) to external "things" (chat/email/etc.)? It doesn't have to be two ways.. just getting information is fine for now..
I find my self setting off multi issue work that is just running on my computer.. I leave the house and just hope its still running.. or am coming constantly back to the computer to see if its finished...
Just getting the following would be awesome:
"Still working on 123. Parker is fixing tests as he has been doing last 10 minutes"
"All work on 123 is done. Team has started on 321"
If I could get periodical information on what is going on in e.g. a Team channel I would be super happy and could be doing something more productive in the meantime like doing the laundry 😅
Triage: Feature request for periodic status updates to external systems (chat/email). Nice-to-have for observability. P2 backlog. Relates to monitoring and notifications — good fit for @telemetry.
Issue:#308 — Periodical monitoring status pushed to external systems?
Telemetry will pick this up in the next Copilot session.
For Copilot coding agent: If enabled, this issue will be worked automatically.
Otherwise, start a Copilot session and say: Telemetry, work on issue #308
Issue:#308 — Periodical monitoring status pushed to external systems?
Telemetry will pick this up in the next Copilot session.
For Copilot coding agent: If enabled, this issue will be worked automatically.
Otherwise, start a Copilot session and say: Telemetry, work on issue #308
Ralph's loop (and squad watch) would POST a JSON payload to the webhook URL whenever something changes. This works with Slack, Teams, Discord, or any webhook-compatible service.
Option B: MCP-based (richer)
If the user has a Teams/Slack MCP server configured, Ralph could post directly to a channel. We already have MCP integration patterns for this.
⚠️ Actions minutes warning
If using the GitHub Actions heartbeat for this, be aware that cron jobs consume Actions minutes. At a 5-minute interval, that's ~8,640 minutes/month. GitHub Free gives 2,000 minutes/month. We should:
Default to a conservative interval (30 min)
Show estimated monthly minutes in the setup flow
Recommend squad watch (local, free) over Actions for frequent polling
Want us to pick this up? The webhook approach is probably the quickest win.
Closing as I think between watch, loop, and the upcoming ideas around dispatch, we've covered the ability for "squad to run in an agentic loop in various ways."
Activity
Merge pull request github#308 from bradygaster/squad/wave1-remaining
chore(.squad): Wave 2 launch — REPL polish, security fix, config extr…
bradygaster commentedon Mar 10, 2026
Triage: Feature request for periodic status updates to external systems (chat/email). Nice-to-have for observability. P2 backlog. Relates to monitoring and notifications — good fit for @telemetry.
github-actions commentedon Mar 10, 2026
📋 Assigned to Telemetry (Aspire & Observability)
Issue: #308 — Periodical monitoring status pushed to external systems?
Telemetry will pick this up in the next Copilot session.
github-actions commentedon Mar 13, 2026
📋 Assigned to Telemetry (Aspire & Observability)
Issue: #308 — Periodical monitoring status pushed to external systems?
Telemetry will pick this up in the next Copilot session.
bradygaster commentedon Mar 22, 2026
Solution: Ralph Loop + External Notifications
We actually have most of the pieces for this already:
What exists today
squad watch(npx @bradygaster/squad-cli watch --interval N) — standalone local process that polls GitHub every N minutes while you're away.squad-heartbeat.yml— GitHub Actions cron for fully unattended monitoring.What's missing: the notification layer
The gap is pushing status updates to external systems. Here's the proposed approach:
Option A: Webhook-based (simplest)
Ralph's loop (and
squad watch) would POST a JSON payload to the webhook URL whenever something changes. This works with Slack, Teams, Discord, or any webhook-compatible service.Option B: MCP-based (richer)
If the user has a Teams/Slack MCP server configured, Ralph could post directly to a channel. We already have MCP integration patterns for this.
If using the GitHub Actions heartbeat for this, be aware that cron jobs consume Actions minutes. At a 5-minute interval, that's ~8,640 minutes/month. GitHub Free gives 2,000 minutes/month. We should:
squad watch(local, free) over Actions for frequent pollingWant us to pick this up? The webhook approach is probably the quickest win.
diberry commentedon Mar 31, 2026
🔄 Ralph status: Feature — external system monitoring push. No PR. Backlog.
bradygaster commentedon Apr 12, 2026
Closing as I think between
watch,loop, and the upcoming ideas arounddispatch, we've covered the ability for "squad to run in an agentic loop in various ways."