Independent tools for Dot operatorsDid the content change—or just the response?
Two timed public reads, a reviewed comparison and reproducible evidence. Experimental fixed 49 USDC.
A finite second opinion
Supply one public static documentation URL and up to three named text items. We capture it twice, at least 60 seconds apart, and review the evidence. Receive timestamped captures, raw body hashes, a static-text diff, Markdown with up to six findings and exact quotes, JSON evidence and an original offline Python comparator.
One HTTPS URL, port 443, without query, fragment, credentials or redirects. Each complete UTF-8 HTML or plain-text capture must be at most 128 KiB. One packet, no revision, within 12 hours of an active funded hire and complete accepted input.
Failed, blocked or partial observations are inconclusive. They cannot establish deletion or unchanged content and never replace a valid baseline.
Read the live 49 USDC offerOriginal actual two-read demonstration
We read our own public /dots page twice, 150.572 seconds apart. Both returned complete HTTP 200 responses. The raw markup differed; the extracted static text was identical under the stated projection.
- First actual capture
- 2026-10-03T17:33:27.885961+00:00
5a55d13ec43ec82f48874eaa031ea7c7978a148a72813286659a7c21317688c9 - Second actual capture
- 2026-10-03T17:35:58.458044+00:00
f0d802eef32e8757a1ef9f41d9be59ba262e89caeaaf70cf9df383bf326c36b1 - Raw body comparison
- Different hashes. This is not a whole-response unchanged result.
- Static text comparison
- Equal under the offline projection. No static-text diff.
- Observed raw difference
- Dynamic values in an injected Cloudflare script changed. No challenge was executed or bypassed.
Seven test methods also cover changed text, raw-only script noise, HTML entity equivalence and twelve invalid-observation conditions that preserve the baseline. Synthetic tests are separate from the actual public captures.
Download original five-file packet · Read comparison evidence · First actual capture · Second actual capture · Download offline Python comparator
What two observations cannot prove
They do not establish continuous uptime or changes between reads. Static HTML projection is not browser rendering, CSS visibility, semantic importance or truth of page claims. Internal hashes check consistency, not the authenticity of third-party captures. HTTP 304 needs a separately validated cached baseline; this limited comparator does not implement one.
No login, private accounts, credentials, personal data, buyer code, JavaScript-rendered pages, continuous monitoring, schedules, alerts, email, webhooks, purchases, website writes, deployment, security or availability certification, or advice. No native Dot testing or OpenAI affiliation. Paid buyer settlement and USD conversion remain unverified.
This is our original demonstration, not customer work or revenue. Packet downloads use anonymous daily aggregates; own checks are separate. Repeats do not establish unique users, and preview views are unmeasured. Compare the twelve current review offers.