diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index 99ee6c5..3cdcc33 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -50,7 +50,7 @@ jobs:
       - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
         with:
           python-version: "3.13"
-      - uses: astral-sh/setup-uv@20cfd1bf945f4377ade1205e4dbc17946fc9a30d # v10.0.1
+      - uses: astral-sh/setup-uv@bec219d24cd3e171d82865faccec33120bb574f4 # v10.1.0
 
       # Keyed on the Makefile because that is where the tool versions are
       # pinned; a bump there must miss the cache and reinstall, or the job would
@@ -204,7 +204,7 @@ jobs:
       - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
         with:
           python-version: "3.13"
-      - uses: astral-sh/setup-uv@20cfd1bf945f4377ade1205e4dbc17946fc9a30d # v10.0.1
+      - uses: astral-sh/setup-uv@bec219d24cd3e171d82865faccec33120bb574f4 # v10.1.0
 
       # perl is on the ubuntu runners, so TestPerlParity runs here rather than
       # skipping -- which makes this the one job that proves the parity fixtures
@@ -224,7 +224,7 @@ jobs:
       - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
         with:
           python-version: "3.13"
-      - uses: astral-sh/setup-uv@20cfd1bf945f4377ade1205e4dbc17946fc9a30d # v10.0.1
+      - uses: astral-sh/setup-uv@bec219d24cd3e171d82865faccec33120bb574f4 # v10.1.0
       - uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
         with:
           path: .venv-tools
@@ -244,7 +244,7 @@ jobs:
       - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
         with:
           python-version: "3.13"
-      - uses: astral-sh/setup-uv@20cfd1bf945f4377ade1205e4dbc17946fc9a30d # v10.0.1
+      - uses: astral-sh/setup-uv@bec219d24cd3e171d82865faccec33120bb574f4 # v10.1.0
       - uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
         with:
           path: .venv-tools
@@ -269,7 +269,7 @@ jobs:
       - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
         with:
           python-version: "3.13"
-      - uses: astral-sh/setup-uv@20cfd1bf945f4377ade1205e4dbc17946fc9a30d # v10.0.1
+      - uses: astral-sh/setup-uv@bec219d24cd3e171d82865faccec33120bb574f4 # v10.1.0
 
       # ccache cuts a rebuild of unchanged C from minutes to seconds, and Nuitka
       # generates a great deal of C that does not change between runs.
diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml
index db7bb8b..2d6e327 100644
--- a/.github/workflows/release.yml
+++ b/.github/workflows/release.yml
@@ -45,7 +45,7 @@ jobs:
       - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
         with:
           python-version: "3.13"
-      - uses: astral-sh/setup-uv@20cfd1bf945f4377ade1205e4dbc17946fc9a30d # v10.0.1
+      - uses: astral-sh/setup-uv@bec219d24cd3e171d82865faccec33120bb574f4 # v10.1.0
 
       # First, and before anything is built: this is the tag-versus-tree check.
       # A `v1.1.0` tag on a commit that still says 1.0.0 would otherwise upload
@@ -85,7 +85,7 @@ jobs:
       - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
         with:
           python-version: "3.13"
-      - uses: astral-sh/setup-uv@20cfd1bf945f4377ade1205e4dbc17946fc9a30d # v10.0.1
+      - uses: astral-sh/setup-uv@bec219d24cd3e171d82865faccec33120bb574f4 # v10.1.0
 
       # No ccache here, deliberately, and the contrast with ci.yml is the point.
       # GitHub's cache is writable from pull-request runs, so a release build