Bunch of new fingerprints

[ghost]

A bunch of updated fingerprints: Chrome #397, Firefox #391, Android #392, Edge (but in more proper and polished form then these PRs; 391 is wrong a little bit, for example), everything tested in wireshark

[SaamoCha]

Thanks for putting this together. I ran an independent review with Parroteer, an experimental tool I built for capturing real browser/Android TLS fingerprints, comparing them against existing uTLS parrots/baselines, and replay-verifying generated specs.

Parroteer can also run in GitHub Actions: it captures up-to-date fingerprints, compares them with the existing fingerprints, and can automatically report detected drift. It is still early, but I hope it can be useful for a future fingerprint auto-update/reporting pipeline. Feedback, usage, and discussion are very welcome.

Target	Verification result
Chrome	trust_anchors support is real, and HelloChrome_141 matches Chrome when TLSTrustAnchorIDs is enabled. However, my local Chrome 149.0.7827.53 default/headless capture did not send trust_anchors and still matched current HelloChrome_133 / HelloChrome_Auto. ML-DSA (0904, 0905, 0906) also looks experimental/non-default in my capture, so HelloChrome_144 seems useful as a selectable profile, but I am less sure it should be Auto.
Firefox	Confirmed. Real Firefox Stable 151.0.3 matches the updated PR shape: JA3N 8099457c290ccfe8c6d958826c26b023, JA4 t13d1617h2_86a278354501_3cbfd9057e0d. Current upstream differs by keeping cipher 49161 and missing extensions 35 / 45.
Android	Confirmed. A real GitHub Actions Android API 36 emulator with OkHttp 5.3.0 matches HelloAndroid_16_OkHttp: JA3N ab22845823d9af0946cb87050d0a5679, JA4 t13d1513h2_8daaf6152771_eca864cca44a. Current HelloAndroid_11_OkHttp is clearly stale.

Android verification run: https://github.com/SaamoCha/parroteer/actions/runs/27101201876

So my read is: Firefox and Android updates are strongly supported by the captures. For Chrome, the new Trust Anchor / ML-DSA profiles are useful, but the Auto choice may need a bit more care because the observed default behavior can vary by feature/field-trial state.

[ghost]

Ok, I think I should keep chrome auto on older version, because these additions are experimental