The animation engine for agents: blazingly fast math and 3D videos with Manim's Python API, on a Rust GPU core.
Python
"""The agent skill (`skills/manimgx/SKILL.md`) is what `npx skills add academa-labs/manimgx`
and `gh skill install academa-labs/manimgx` install: they find it by its path,
`skills/<name>/SKILL.md`, and read its front matter, which the Agent Skills specification
(https://agentskills.io/specification) constrains. This checks it against the specification:
the name is its folder's, and the description is 1 to 1,024 characters. Its authoring
examples also run against the current API."""
import re
from pathlib import Path
import yaml
import manimgx as m
SKILL = Path(__file__).parents[2] / "skills" / "manimgx" / "SKILL.md"
def test_its_front_matter_follows_the_specification() -> None:
text = SKILL.read_text(encoding="utf-8")
match = re.fullmatch(r"---\n(.*?)\n---\n.*", text, re.DOTALL)
assert match, "SKILL.md opens with its front matter, between two `---` lines"
fields = yaml.safe_load(match[1])
assert fields["name"] == SKILL.parent.name
assert isinstance(fields["description"], str)
assert 1 <= len(fields["description"]) <= 1024
def test_its_authoring_examples_run() -> None:
blocks = re.findall(
r"^```python\n(.*?)^```",
SKILL.read_text(encoding="utf-8"),
re.MULTILINE | re.DOTALL,
)
assert blocks, "the authoring skill needs an executable scene example"
for code in blocks:
namespace: dict[str, object] = {}
exec(compile(code, str(SKILL), "exec"), namespace)
scenes = [
value
for value in namespace.values()
if isinstance(value, type)
and issubclass(value, m.Scene)
and value is not m.Scene
]
assert scenes, code
for scene in scenes:
scene().render() # Evaluate and count frames without drawing a video.