Summarize Incidents and On-Call Into a Daily Note
By the end of this recipe, you'll get a clear reliability note that turns your incidents board into a short daily read — what's open and how bad, what breached SLA, what resolved overnight, and who's on call today — either on demand from Ask Copera AI or posted automatically every morning. Because your incidents, your chat, your runbooks, and your on-call rota all live in one workspace, the note can pull the whole picture together instead of making someone stitch it from five tabs at standup.
Readiness: Available now — ask Copera AI to write the reliability note on demand, or build the scheduled digest as an Automation. The version where a named teammate posts it every morning on its own is rolling out in early access.
What it does
Reliability status usually lives in someone's head until standup, then gets re-explained in a scramble: which incidents are still open, which one is about to breach, what closed overnight, who's holding the pager today. This recipe answers all of that from the record itself.
The reliability note is a short, plain-language read-out — not a data dump — that covers:
| The note covers | Pulled from |
|---|---|
| Open incidents — severity, owner, age, and current status | The incidents board |
| SLA & response breaches — anything past its goal clock, or close to it | The SLA timers on the board |
| Resolved in the last 24 hours — what closed and how long it took | The incidents board |
| On call today — who's holding the pager | Your on-call rota (a board or a schedule) |
| Watch items — incidents going quiet, reopening, or piling on one owner | The board's activity |
There are two honest ways to produce it, and this page gives you both:
- On demand — ask Copera AI for the note whenever you want it (before standup, at an on-call handoff, when a VP asks "how's reliability?"). This is the fullest version, because the assistant can read across the board, your
#incidentschat, your runbook docs, and the rota in one pass. - Automatically — a scheduled automation on the incidents board writes the note every morning and posts it, hands-off. It's scoped to that one board's fields, which is the reliable thing for an automation to own.
Build it yourself
A scheduled automation reads one incidents board each morning and narrates it. That's the honest, reliable scope for an automation — it's attached to a board and reports on that board's fields. For a note that also weaves in your runbook docs, related chat, and the on-call rota, use the on-demand ask below (or a standing teammate).
Set up an Incidents board (or reuse your existing one) with columns for Severity (Sev-1/2/3), Status (Investigating, Mitigated, Resolved…), Owner, an Opened date, and a Countdown SLA timer for response or resolution time. These are the fields the note is built from, so the richer the board, the sharper the read. See Boards Overview if you're starting fresh.
In Automations, add a Schedule trigger for each morning — for example, 8:00 AM, Monday through Friday, timed just before your team logs on. On-call teams often run a second copy at shift-change time as a handoff note.
Add a Find rows action that returns the incidents worth reporting: everything not yet Resolved, plus anything resolved in the last 24 hours, so the note covers both what's still burning and what just closed. A row filter keeps it to the severities and statuses you care about.
Add an AI → Summarize step over those rows. Prompt it plainly: "Write a short daily reliability note. Lead with a one-line headline (how many open, worst severity). List each open incident with its severity, owner, age, and status. Call out anything past or near its SLA. Then a line on what resolved in the last 24 hours. Keep it tight." Swap in AI → Generate if you want it to follow a fixed template every day.
Add a Post channel message action to your #reliability or #eng channel so the whole team sees it. Prefer a permanent record? Add a Create document action to append the day's note to a running reliability log. Reporting up? Add a Send or draft email action to leadership — use draft if you want to top-and-tail it first.
Add real-time alerts, not just the daily recap. The digest is the morning read; pair it with a second automation that fires the instant something happens. Trigger on Row created (a new incident is logged) or Field changed (severity raised to Sev-1) and add a Post channel message plus a Send direct message to the current on-call owner — so a Sev-1 at 2 a.m. reaches a person immediately, and the morning note simply confirms what everyone already saw. For breaches specifically, use the SLA-Breach Watchdog pattern with an SLA breached trigger.
An automation can't read across your whole workspace — it narrates the board it's attached to. That's deliberate and dependable. When you want the note to also reflect your runbook docs, the #incidents conversation, and who's actually on call, that cross-surface reading is what Ask Copera AI does on demand (next), and what a standing teammate will do on a schedule (early access).
Or just ask Copera AI
This is the fullest version of the reliability note, and the fastest way to get one right now — no build required. Open Ask Copera AI and ask:
"Read our Incidents board and write today's reliability note: every open incident with its severity, owner, age, and current status; anything that has breached or is close to breaching its SLA; and what resolved in the last 24 hours. Check the on-call rota and name who's on call today, pull in anything relevant from the #incidents channel and our runbook docs, and post the note to the #reliability channel."
Ask Copera AI reads the incidents board, searches the workspace for the related discussion, reads the runbook docs you point it at, checks the on-call schedule, and writes the note — then, because your data is unified, it can post it to the channel in the same turn. It answers with a Sources list, so you can jump straight to any incident or message it cited. With the default Ask before acting permission mode, it shows a review card before it posts anything, so you approve the wording before it goes out. See What Copera AI Can Do for the full list of actions it can take.
It's just as good for the one-off questions that come up all day: "What's the status of the checkout outage, and who's on it?" or "Summarize everything that broke this week for the postmortem." You read the board's whole story back in a sentence instead of scrolling it.
Make it a standing teammate
AI teammates are rolling out gradually. If you don't see them in your workspace yet, they're on the way — everything in this guide can be built today as an Automation. When one is available, a named Reliability teammate can own this end to end: read the incidents board, the on-call rota, and the day's #incidents chat each morning, write the reliability note, and post it — plus keep the on-call handoff current — on its own schedule and under its own identity. Until then, the scheduled automation and the on-demand ask above do the job.
Tips
Lead with the exception, keep the all-clear to one line. A note that lists every incident at equal weight gets skimmed and ignored. Prompt the AI to open with the headline — "2 open, 1 Sev-1 breaching in 20 min" — and put the calm details below. On a quiet day, "No open incidents; on call: Priya" is the whole note, and that's a good day.
Make the handoff a first-class use. On-call changes hands on a schedule, so run a second copy of the digest at shift-change time — or just ask Copera AI "write the on-call handoff: what's open, what to watch tonight, and where the runbooks are." The incoming engineer inherits context instead of a cold pager.
Feed postmortems from the same board. Because every incident is a row with an owner, timeline, and SLA history, ask Copera AI to "draft the postmortem for the checkout outage from its row and comments" — or turn a review meeting into follow-up tasks with the Meeting-to-Tasks recipe. The daily note and the retro run off one record.
Set a spend cap on the digest. The morning automation runs an AI summary every day; give it a per-automation spend cap and watch the usage dashboard so a week of noisy incidents can never run past a budget you choose — the run simply stops and reports it.
Frequently Asked Questions
Can Copera AI summarize my incidents board?
Yes. Ask Copera AI reads your incidents board on demand and writes a reliability note — every open incident with its severity, owner, age, and status; anything that breached its SLA; and what resolved in the last day. Because your data is unified, it can also fold in related chat, your runbook docs, and who's on call, then post the note to a channel or save it as a document.
Can I get an automatic daily incident report?
Yes. Build a scheduled automation on the incidents board: it runs each morning, uses a Find rows step to gather the open and recently resolved incidents, an AI Summarize step to write the note, and a Post channel message or Create document action to deliver it. No one has to write the report — it's waiting when the team logs on.
Is Copera an incident management or on-call tool?
Copera isn't a dedicated pager, but teams run lightweight incident tracking on a board — severity, owner, status, and SLA timers — and Copera's AI and automations turn that board into real-time alerts and a daily reliability note. You keep the incident record where the rest of your work already lives, and the intelligence layer narrates it for you.
Can an automation alert the on-call engineer when a new incident opens?
Yes. A Row created or Field changed trigger fires the moment an incident is logged or escalated, and a Post channel message plus Send direct message action alerts the current on-call owner and your incident channel instantly — no waiting for the daily digest.
Can Copera escalate when an incident breaches its response time?
Yes. Put a Countdown SLA timer on each incident and add an SLA breached trigger. The moment the response or resolution clock passes its goal, the automation flags the incident and escalates it to the on-call lead — the same pattern as the SLA-Breach Watchdog.
Where does the reliability note get posted?
Wherever it's useful. Post it to a #reliability or #eng channel for the team, append it to a running document as a permanent log, or email it to leadership — it's the same note with a different destination. Ask Copera AI can send it in the same turn it writes it, or the automation can deliver it automatically.
Related
- Automations — the schedule trigger, find-rows, AI summarize, and messaging blocks this recipe is built from
- Ask Copera AI — write the reliability note, or answer incident questions, on demand
- What Copera AI Can Do — the full list of actions the assistant can take
- SLA-Breach Watchdog — the real-time escalation cousin, for the moment an incident breaches
- An AI End-of-Day Team Check-In — the same daily-digest pattern for team status
- What You Can Build with Copera AI — the full playbook