Identity / Brute force · Notable alert
Repeated failed sign-ins for one account
A rule has fired because one account on a standard workstation failed to sign in many times in a few minutes, all from the same outside address. The workstation's owner has already called the help desk to say he cannot sign in. No earlier alert for that address is on the queue.
How a careful person works it
- Read what the rule says it saw. The alert names the account, the address, the machine and the time window. Everything else you check is measured against those.
- Count and compare in the logs. One account hit many times is a different pattern from many accounts hit a few times, and the numbers tell you which one you have.
- Look for a success after the failures. Failed attempts alone are noise on the internet. A success from the same source would make this a possible compromise.
- Decide, and say it as a verdict. True positive, false positive or needs more data are the only honest options, and the choice drives the response.
- Act at your level and hand over the facts. A first-level analyst blocks the source and passes the case on with the evidence, instead of acting beyond their authority.
The grader reads what you pinned as evidence, the query you ran, the action you took and the case note. It rewards a correct verdict reached from the data, not a lucky guess, and it flags a response that went further than the evidence supported.