refactor: unify C2PA vendor registry + code-health fixes + uv publish

Three P2 cleanups from a library-wide review.

Detection -- single C2PA_AI_VENDORS registry (noai/constants.py):
- C2PA_ISSUERS, SYNTHID_C2PA_ISSUERS, and identify._ISSUER_PLATFORM now derive
  from one C2paAiVendor table, so adding a C2PA vendor is one entry instead of
  edits in three places across two files. Behavior-identical (262 detection
  tests pass; the kept `needle` field is load-bearing -- it differs from `org`
  for Google and ByteDance, with no mechanical derivation).

Code-health:
- region_eraser.erase_lama now accepts grayscale/BGRA like erase_cv2 (it
  crashed on grayscale and silently dropped alpha on BGRA). +2 regression tests.
- batch frees the device cache between images via a shared try_empty_device_cache
  helper (generalized from the MPS-only _try_clear_mps_cache, now reused by both
  the MPS->CPU fallback and the batch loop).
- batch gained --controlnet-scale (parity with invisible/all).

CI / packaging:
- publish.yml uploads via `uv publish` (PyPI trusted publishing over OIDC),
  replacing pypa/gh-action-pypi-publish so uploads no longer depend on that
  action's bundled twine accepting the Metadata-Version. Workflow filename +
  pypi environment unchanged, so PyPI's trusted-publisher entry still matches.
- hatchling pin relaxed <1.28 -> <1.31 (verified against hatch's changelog:
  1.30.0 made Metadata 2.5 the default, 1.30.1 reverted to 2.4; 1.27-1.29 were
  always 2.4). Kept as belt-and-suspenders so the first uv-publish release ships
  2.4, isolating the uploader swap from the metadata-version bump.

Docs (CLAUDE.md, pyproject) synced; corrected the inaccurate "hatchling 1.28+
emits 2.5" note.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

tests/test_region_eraser.py

@@ -90,3 +90,47 @@ class TestLamaBackend:
             pytest.skip("onnxruntime installed; cannot test the unavailable path")
         with pytest.raises(RuntimeError, match="onnxruntime"):
             erase(img, boxes=[(10, 10, 20, 20)], backend="lama")
+
+
+class TestLamaChannelHandling:
+    """erase_lama must accept grayscale (2D) and BGRA (4-channel) like erase_cv2.
+
+    The real ONNX model is never loaded -- the session is faked to an identity
+    inpaint, so this exercises only the channel promote/split wrapper (the fix for
+    LaMa crashing on grayscale and dropping alpha on BGRA).
+    """
+
+    @pytest.fixture
+    def _fake_lama(self, monkeypatch: pytest.MonkeyPatch):
+        from remove_ai_watermarks import region_eraser
+
+        class _In:
+            def __init__(self, name: str, shape: list[int]):
+                self.name = name
+                self.shape = shape
+
+        class _FakeSession:
+            def get_inputs(self):
+                return [_In("image", [1, 3, 512, 512]), _In("mask", [1, 1, 512, 512])]
+
+            def run(self, _outputs, feeds):
+                # Identity inpaint: echo the image tensor (1,3,size,size) back.
+                return [feeds["image"]]
+
+        monkeypatch.setattr(region_eraser, "lama_available", lambda: True)
+        monkeypatch.setattr(region_eraser, "_get_lama_session", lambda: _FakeSession())
+
+    @pytest.mark.usefixtures("_fake_lama")
+    def test_grayscale_2d_does_not_raise(self):
+        gray = np.full((100, 100), 120, np.uint8)
+        out = erase(gray, boxes=[(40, 40, 20, 20)], backend="lama")
+        assert out.ndim == 2
+        assert out.shape == gray.shape
+
+    @pytest.mark.usefixtures("_fake_lama")
+    def test_bgra_preserves_alpha(self):
+        bgra = np.full((100, 100, 4), 120, np.uint8)
+        bgra[..., 3] = 200  # opaque-ish alpha plane
+        out = erase(bgra, boxes=[(40, 40, 20, 20)], backend="lama")
+        assert out.shape == bgra.shape
+        assert np.array_equal(out[..., 3], bgra[..., 3])  # alpha carried through unchanged