| name | blender-tests |
| description | Run this repo's headless Blender test suite (tests/run_tests.py) via podman against one or more pinned Blender versions. Use before a commit/PR, or to check cross-version compatibility. |
Run the headless Blender test suite
A fixed script instead of an ad-hoc podman run ... -v "$(pwd)":/repo:Z ... invocation, so it can
be whitelisted in permission settings by exact path rather than needing a broader podman run *
rule (the $(pwd)-based mount is what made the ad-hoc version unsafe to whitelist as a fixed
pattern — every invocation embeds a dynamic shell substitution).
.claude/skills/blender-tests/run_blender_tests.sh <version> [<version> ...]
# e.g.
.claude/skills/blender-tests/run_blender_tests.sh 5.2
.claude/skills/blender-tests/run_blender_tests.sh 4.1 4.5 5.0 5.1 5.2
Pulls docker.io/blenderkit/headless-blender:blender-<version>-stable if not already present,
mounts the repo read-write at /repo inside the container, and runs
blender --background --python-exit-code 1 --python /repo/tests/run_tests.py. Exit 0 only if
every version's test run exits 0. Doesn't rely on the caller's $(pwd) or take a repo path — it
derives the repo root from its own on-disk location, so call it directly from any cwd.
Matches this repo's CI matrix (currently 4.1/5.2, see .github/workflows/ci.yml) — use the
full 4.1 4.5 5.0 5.1 5.2 set for a broader local check before a version-support-affecting
change, matching what CLAUDE.md's "Known Blender version support" section says has been
spot-checked.