text
| 1 | # Status Update Template |
| 2 | |
| 3 | Post at the cadence implied by severity (see `WORKFLOW.md` Phase 3). Keep it factual |
| 4 | and brief — this is read by people who need the current state, not the investigation |
| 5 | narrative. |
| 6 | |
| 7 | ```markdown |
| 8 | **[<severity>] <one-line incident title>** — <INVESTIGATING | MITIGATING | MONITORING | RESOLVED> |
| 9 | |
| 10 | **Impact:** <what's broken, for whom, since when — concrete, not vague> |
| 11 | **Current status:** <what's known right now, in plain language> |
| 12 | **Actions in progress:** <what's actively being done> |
| 13 | **Next update:** <time, per the severity cadence> |
| 14 | ``` |
| 15 | |
| 16 | ### Resolved update |
| 17 | |
| 18 | ```markdown |
| 19 | **[<severity>] <title>** — RESOLVED |
| 20 | |
| 21 | **Resolution:** <what fixed it, when impact actually stopped (verified, not assumed)> |
| 22 | **Duration:** <start time> to <resolution time> |
| 23 | **Postmortem:** <link, or "to follow within 1 business day"> |
| 24 | ``` |
| 25 | |
| 26 | ## Guidance |
| 27 | |
| 28 | - Never speculate on root cause in a status update before it's confirmed — "we believe |
| 29 | this may be related to X, still confirming" is fine; presenting a guess as fact is |
| 30 | not. |
| 31 | - If severity is re-classified mid-incident, say so explicitly in the next update |
| 32 | ("upgrading this to sev1 — impact is broader than initially assessed") rather than |
| 33 | silently changing the cadence. |
| 34 | - A resolved update is not optional — always post one, even if it's brief. |
| 35 |