Fix Windows ORT builds and Docker signing retries

.github/workflows/docker.yml

@@ -331,7 +331,18 @@ jobs:
         run: |
           target="${IMAGE}@${INDEX_DIGEST}"
           echo "Signing ${target} (signatures -> ${COSIGN_REPOSITORY})"
-          cosign sign --yes "${target}"
+          for attempt in 1 2 3; do
+            if cosign sign --yes "${target}"; then
+              exit 0
+            fi
+            if [ "$attempt" -eq 3 ]; then
+              echo "ERROR: cosign signing failed after ${attempt} attempts" >&2
+              exit 1
+            fi
+            sleep_for=$((attempt * 10))
+            echo "cosign signing failed on attempt ${attempt}; retrying in ${sleep_for}s" >&2
+            sleep "$sleep_for"
+          done
 
   promote-latest:
     # Re-push the :latest tag pointing at the root variant *after* every

Cargo.lock

@@ -2487,6 +2487,16 @@ dependencies = [
  "cc",
 ]
 
+[[package]]
+name = "libloading"
+version = "0.9.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "754ca22de805bb5744484a5b151a9e1a8e837d5dc232c2d7d8c2e3492edc8b60"
+dependencies = [
+ "cfg-if",
+ "windows-link",
+]
+
 [[package]]
 name = "libredox"
 version = "0.1.16"
@@ -2926,6 +2936,7 @@ version = "2.0.0-rc.12"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d7de3af33d24a745ffb8fab904b13478438d1cd52868e6f17735ef6e1f8bf133"
 dependencies = [
+ "libloading",
  "ndarray",
  "ort-sys",
  "smallvec",

crates/headroom-core/Cargo.toml

@@ -61,11 +61,6 @@ flate2 = "1"
 # (release matrix, e2e/wrap, e2e/init, devcontainer, ci.yml) needs
 # system OpenSSL + perl modules for openssl-src vendored compile. Using
 # rustls everywhere removes the entire OpenSSL build-deps surface.
-fastembed = { version = "5", default-features = false, features = [
-    "hf-hub-rustls-tls",
-    "ort-download-binaries-rustls-tls",
-    "image-models",
-] }
 # `magika` is Google's ONNX-backed content classifier (Tier 1 of the
 # Stage-3d ContentRouter detection arch). Bundled standard-model is
 # loaded once per process via `OnceLock` and shared across calls. The
@@ -127,6 +122,23 @@ redis = { version = "0.27", optional = true, default-features = false }
 # cycling through the proxy crate. Tiny crate (no I/O, just types).
 http = "1"
 
+[target.'cfg(not(windows))'.dependencies]
+fastembed = { version = "5", default-features = false, features = [
+    "hf-hub-rustls-tls",
+    "ort-download-binaries-rustls-tls",
+    "image-models",
+] }
+
+[target.'cfg(windows)'.dependencies]
+# `ort-download-binaries-*` emits DirectML link libs on Windows (`DXCORE`,
+# `DXGI`, `D3D12`, `DirectML`). Users installing `headroom-ai[all]` from
+# sdist often do not have those SDK libs, so load ORT dynamically instead.
+fastembed = { version = "5", default-features = false, features = [
+    "hf-hub-rustls-tls",
+    "ort-load-dynamic",
+    "image-models",
+] }
+
 [features]
 default = []
 # Compile in the Redis CCR backend. Enable for multi-worker deployments

tests/test_release_workflows.py

@@ -187,6 +187,27 @@ def test_fastembed_uses_rustls_features() -> None:
     assert '"image-models"' in cargo
 
 
+def test_fastembed_uses_dynamic_ort_on_windows() -> None:
+    """Windows sdist builds must not link Pyke's DirectML ORT binaries.
+
+    `ort-download-binaries-*` emits DXCORE/DXGI/D3D12/DirectML link libs on
+    Windows. Those SDK libs are not present on many Python build hosts, so the
+    Windows target must use ORT dynamic loading instead.
+    """
+
+    cargo = (ROOT / "crates" / "headroom-core" / "Cargo.toml").read_text(encoding="utf-8")
+    assert "[target.'cfg(windows)'.dependencies]" in cargo
+    windows_section = cargo.split("[target.'cfg(windows)'.dependencies]", 1)[1].split(
+        "\n[",
+        1,
+    )[0]
+    windows_dependency_lines = "\n".join(
+        line for line in windows_section.splitlines() if not line.lstrip().startswith("#")
+    )
+    assert '"ort-load-dynamic"' in windows_section
+    assert "ort-download-binaries" not in windows_dependency_lines
+
+
 def test_dockerfiles_no_longer_install_openssl_devel() -> None:
     """Once openssl-sys is out of the build tree, every Dockerfile
     that used to install `openssl-devel` / `libssl-dev` for the Rust