Merge origin/main into docs/library-research-landscape

Main's claim_generator_platform and c2pa_vendor_platform now use the
branch's whole-word registry match, which replaces
claim_generator_word_platform. The branch's provenance veto covers every
video mark and made main's Veo-over-Sora override unreachable, so the
override is removed. Main's credential rule carries into the veto: an
invalid credential vetoes nothing. Provenance that confirms the mark's
own vendor also vetoes nothing. Over 78 local videos with a C2PA AI
claim, the only verdict changes against main are Kling/Hailuo false
positives on Veo and FLUX clips that the veto now removes. C2PA verdicts
over 1,379 files are unchanged against the pre-merge branch.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

scripts/render_visible_examples.py

@@ -267,9 +267,8 @@ def _video_mark_frame(key: str, w: int, h: int) -> np.ndarray:
     if key == "sora":
         tmpl, scale, x, y = templates["sora-icon"], 0.10, int(w * 0.72), int(h * 0.90)
     elif key == "veo":
-        # The legacy "Veo" TEXT form: a perfect synthetic diamond also matches the
-        # Sora icon template (both are 4-point stars) and table order hands the
-        # tie to Sora, so the gallery carries the discriminative text variant.
+        # Keep the legacy text variant in the generated gallery. The provider
+        # original and the cross-match regression exercise the current diamond.
         tmpl = templates["veo-text"]
         th = max(6, round(14 * short / 720))
         tw = max(1, round(tmpl.shape[1] * th / tmpl.shape[0]))