Incident Comms Timeline Logger
Log every beat of an incident - detection, internal actions, each customer message, and resolution - then let AI draft a post-incident summary your review approves before it becomes the official record.
A private team tool where you log a running incident timeline (typed in or imported from your broadcaster/suppressor logs), get an AI-drafted chronological summary and customer-impact recap, edit and approve it, then export and email an official post-incident report to stakeholders.
Before you start
- A free Supabase account
- A free Resend account (or use the CSV fallback to start)
- A Vercel account for deployment
- Your last incident's notes or chat log to test with
The problem this kills
When an incident is over, the review starts - and everyone reconstructs the timeline from memory, scattered chat threads, and screenshots. Who got notified first? When did the status page go up? How long were customers actually affected? The story drifts, blame creeps in, and the "lessons learned" are built on a shaky version of events.
You need one place where the facts land as they happen, in order, with real timestamps - so the post-incident review argues about what to fix, not about what happened.
What you'll build
A private, login-only web app for your incident managers and ops leads. During an incident you log entries - detection, internal actions, and every customer communication - and you can import the messages your other tools already recorded (the broadcaster that sent updates, the suppressor that held them back). When it's over, the AI reads the whole ordered timeline and drafts a clean chronological summary plus a customer-impact recap. You edit it, you approve it, and only then is the official post-incident report exported and emailed to stakeholders.
What's inside the Implementation Plan
The plan opens by interviewing you about your business - your current incident process, the tools that log your comms, how you name and number incidents, your severity levels, your typical and worst-case timelines, and your messy edge cases - so the tool is tailored to how you actually run incidents, not a generic template. Then, step by step with copy-paste prompts:
- A timeline data model shaped to your incident IDs, severities, and entry types.
- A fast logging screen for typed entries with accurate, ordered timestamps.
- An importer that pulls comms from your broadcaster/suppressor logs (or a CSV) and dedupes on the incident ID so nothing lands twice.
- An AI drafter that turns the ordered timeline into a chronological summary and a customer-impact recap.
- A review-and-approve gate, then a one-click export and stakeholder email.
The governance it includes (this is the point)
- Login so only your team can open it.
- Row-level security so each organization only ever sees its own incidents.
- A full audit trail - who logged, edited, and approved each entry, and when.
- A hard human approval gate - the AI only drafts; the incident manager reviews and approves before anything becomes the official record.
- Duplicate guards keyed on the incident ID so the same imported message can't be processed twice.
Who it's for
Incident managers and ops/BPM leads who run post-incident reviews and are tired of rebuilding the timeline from memory every time.
You've got this - paste the first prompt and let the agent interview you.