I don't. The script captures before and after, not during. The gap remains. What I've added is a screen recording (OBS, triggered by hotkey) for the actual purchase flow. The recording + before/after screenshots + confirmation email = as close to continuous as I've gotten.
Still not perfect. The perfect solution would be browser-level transaction attestation, which doesn't exist for consumer use.
The perfect solution is not buying from providers where you need this level of paranoia. But we all know that's not how this market works.
I was productive before Kubernetes and I'm productive before automated purchase-documentation scripts. A simple folder and discipline is enough. The rest is anxiety engineering.
Anxiety engineering describes most of this thread. I have one small server. One provider I've used for 6 years. No disputes, no documentation beyond the invoice.
Not everyone needs 40 VPS and a Git repository of evidence. The question was "what do you use," not "how do I become a forensic accountant for $5/month hosting."
Member
I do though. That's the point. Different scales need different solutions. My "collection of cheap VPS and bad decisions" requires tracking because cheap providers turn into bad decisions regularly.
For my one production server? Different provider, different trust level, lighter documentation. Context matters.
Context does matter. I'm following because I manage hosting for 3 family businesses. Not my job, not my expertise, but I'm the one who "knows computers." When something breaks, I'm the one who gets the call.
Documentation is for them, not for disputes. It's so I can remember what we bought and when it renews. The dispute angle is secondary.
Same here. I "know computers" for my extended family. I need documentation so I can answer "when does the website expire" without panic-searching my email.
Disputes are rare. Confusion is constant.
This thread drifted from "how to document for disputes" to "how to remember what you bought" to "don't buy bad providers." All valid, but the original question about chain of custody at transaction moment is still unanswered.
Is there really no better solution than screenshots + archives + hope?
Because there is no perfect answer. The legal system moves slow, digital evidence is malleable, and hosting contracts are low-value. The effort to prove malice often exceeds the recovery.
Best practical answer: credit card, good monitoring with public logs, and enough documentation to make a chargeback or small claim plausible. Anything more is diminishing returns.
I use Trustpilot reviews as supplementary evidence. If a provider promised me 99.9% and I can show 10 other people reporting the same "your monitoring is wrong" response, it strengthens my case.
Not primary evidence. Context evidence. Helped me with a PayPal dispute once.
Trustpilot is gamed to hell. Providers buy reviews, competitors trash them. I wouldn't rely on it for anything.
For actual monitoring evidence: I run UptimeRobot + HetrixTools + a self-hosted Uptime Kuma instance. Three independent sources. When all three agree, the provider can't claim "your monitoring is wrong" without claiming three different services are wrong simultaneously.
Expensive in time, not money.
