Mode d’emploi
- Copiez la commande et examinez le paquet avant de l’installer dans votre environnement agent.
- Consultez le document original pour les outils, identifiants et environnements requis.
- Décrivez les entrées et le livrable souhaité ; comparez la sortie aux sources.
npx skills add https://github.com/google-deepmind/robotics-antigravity-plugin --skill share-exampleAvant de commencer
- L’installation n’inclut ni abonnement au modèle, ni quota API, ni accès aux fichiers.
- Étoiles et installations reflètent l’activité, pas un test indépendant de qualité ou de sécurité.
Rendu depuis le dépôt source en conservant titres, exemples, code, tableaux, liens et images.
Share Robotics Example to GitHub Discussions
Packages a real-robot task or MuJoCo simulation demo into two drag-and-drop files (demo.mp4 and bundle.zip) and a pre-filled Show and tell URL for https://github.com/google-deepmind/robotics-antigravity-plugin/discussions. Shares only to GitHub Discussions. Never creates a GitHub repo or requires gh.
Steps
- Gather inputs:
- Compact transcript:
<appDataDir>/brain/<conversation-id>/.system_generated/logs/transcript.jsonl
- Final workspace files (controller, robot/sim scripts,
scene.xml, etc.),
skipping .env, .git, .venv, __pycache__, node_modules, and *.mcap.
- Demo videos (
.mp4,.mov, or.giffrom real robot or simulation). - Short title (e.g.
[Real SO-101] Pick Markeror[Sim] YAM Espresso),
prompt used, and setup (real robot model or sim-only + OS).
- Scrub and package into `./share_post/`:
- Copy the compact
transcript.jsonland final text files, redacting API
keys/tokens (AIza..., GEMINI_API_KEY, ghp_..., sk-..., Bearer) and replacing local home paths (/Users/<name>, /home/<name>) with ~.
- Write a short
README.md(title, setup, prompt, file list) and zip
README.md, scrubbed transcript.jsonl, and final files into ./share_post/bundle.zip (must be under 25 MB).
- Copy demo video(s) into
./share_post/. If an.mp4/.movexceeds 10 MB
and ffmpeg is installed, compress it to H.264 (yuv420p) under 10 MB.
- Review and approve (mandatory):
Show the user the files inside bundle.zip, redaction counts, video sizes (remind them to check videos for private lab details), and the draft Discussion title and body. Wait for explicit approval.
- Post to Show and tell:
Build a URL-encoded pre-filled link to https://github.com/google-deepmind/robotics-antigravity-plugin/discussions/new with category=show-and-tell, title, and body (Demo, Prompt, Setup, Files & Conversation). Give the user the link and ./share_post/ so they can open it, drag the video(s) and bundle.zip into the editor, and click Start discussion.