Kastell
A crisis management cockpit that takes over from the information system when you can no longer trust it.
It prepares the crisis in peacetime, runs it under pressure, and records it in full — so that it can be replayed.
What this product assumes
That on the day you need it, your email, your directory and your file shares may be down — or worse: up, but compromised. Everything that follows comes from that assumption.
Three phases, three uses
| Phase | What Kastell does |
|---|---|
| Before — 99 % of the time | Holds the directory, the third parties to call, the response sheets and the document kit, outside your IS. Measures how stale all of it is, and names what is missing. |
| During — a few days | Chat, video, documents, whiteboard, incident log, decisions, actions, situation reports, call-out, continuity, regulatory deadlines. |
| After — years | The log cannot be edited. You replay the crisis at any moment, and a third party can verify the record without running our code. |
What sets it apart
- An append-only log, chained by hashes. The constraint is held by the database: updates and deletes are refused, including to the database owner.
- An independent verifier, dependency-free, released under CC0. A guarantee you can only check on the terms of whoever claims it guarantees nothing.
- Self-hostable, under AGPL-3.0. One Dockerfile, one compose file, nothing crippled. The same artifact serves the hosted offer and your own installation.
- No connection to your corporate directory. Deliberately: that directory may be precisely what is no longer available or trustworthy.
Where to start
- Why this exists — the reasoning, in five minutes.
- Installation — a running instance, in ten minutes.
- How it works — the black box, first.