Supplied allegation
Phia's extension allegedly used an invisible browser tab to plant its tracking code on purchases it had nothing to do with, creating affiliate commissions.
CLAIM AUDIT / PHIA
What is alleged, what is shown, what is still missing.
A calm evidence ledger for a charged post about affiliate commissions and browser-extension tracking. The source is preserved exactly as supplied.
Current posture72%High attention, incomplete proof
Read the claim in layers.
Named corroboration
Bloomberg, a fraud researcher, and a competitor are named as testers.
Known boundary
The supplied post ends mid-sentence and gives no methods, dates, logs, response, or URL.
Verification queue
Obtain the full report, reproduce in a clean profile, inspect requests, and compare attribution timestamps.
INTERACTIVE EVIDENCE POSTURE
How much attention should this claim receive?
Adjust confidence in the supplied corroboration. This changes the posture, never the underlying facts.
High attention
Several named testers are described, but the source is incomplete and methods are absent.
Manual evidence-posture thresholds and an unverified source boundary
Read the explanation
The source has one manually selected confidence value from zero to one hundred, initially seventy two. Bars compare seventy two with zero on a shared five pixel per point scale. The visible percentage and headline posture copy follow that input; they are not calculated from logs, a statistical model or a verified fraud probability. The supplied allegation remains an allegation. Changing the dial does not establish intent, wrongdoing, company behavior or the accuracy of named corroboration. The source label is Low confidence below forty, Watch closely from forty through sixty nine and High attention at seventy and above. Bars show thirty nine, forty, sixty nine and seventy at five pixels per point. Those changes select fixed wording, while the underlying source text and missing evidence remain unchanged. The dial supplies no investigation, network reproduction or external verification; crossing seventy does not prove the allegation. The actual brief exports eight fields, including exact supplied source text, manual confidence, posture, named corroboration, missing evidence and next steps. Bars compare eight fields and four listed missing categories at fifty pixels per item. Those gaps are the full source URL, test methods, network logs and a response. The three suggested next steps are written questions, not completed investigations. Reload restores the manual dial and there is no Reset control. Download produces a local file without accessing a browser extension, testing tracking, contacting anyone or invoking a model.