Go2 Workspace Testing
SkillDev toolsTest and generate proof for go2_ws. Use for workspace build and runtime checks, Isaac Sim joint-command motion, or Isaac Lab rough-terrain training and playback.
Use Go2 Workspace Testing in Claude, ChatGPT or Ahel Desktop
Free. Sign in, add Go2 Workspace Testing and connect your AI. About a minute.
Also: Claude Code · Cursor · Codex
Then ask your AI: use the Go2 Workspace Testing skill
Details
Instructions available. Your AI can read the instructions. Execution depends on the setup they require.
Account requirements not reviewed. Check the skill instructions before use; Ahel provides instructions and does not run this skill.
No other account needed.
Add Ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.
What this skill tells your AI
The instructions your AI receives, as published by j3soon/ros2-essentials in skills/go2-workspace-testing/SKILL.md and read by Ahel’s review.
Build
python3 tests/workspace_smoke/run.py --workspace go2_ws --level build
CLI
python3 tests/workspace_smoke/run.py --workspace go2_ws --level image-cli
python3 tests/workspace_smoke/run.py --workspace go2_ws --level cli
GUI Proof
Runs the documented Custom Isaac Sim Environment demo. The runner opens
go2_og.usda with the Isaac timeline paused, validates /World/go2, starts
host X11 recording, triggers Isaac timeline Play, verifies ROS traffic with
ros2 topic echo --once /clock and ros2 topic echo --once /joint_states,
waits for the scene to stabilize, publishes the documented /joint_command,
then captures a final host X11 screenshot. The MP4 must include the start of
simulation and pre-command frames.
The doc-demo runner builds locally and refuses registry pulls. After a
successful build, add --no-build to reuse the tested image.
python3 tests/workspace_smoke/doc_demo_smoke.py --workspace go2_ws \
--summary-json tests/workspace_smoke/artifacts/go2_ws-doc-demo.json
Report back with the summary JSON path, log path, screenshot path, and MP4 recording path.
Isaac Lab Rough-Terrain Proof
The Go2 Isaac Lab check runs two RSL-RL training iterations, verifies and
copies model_1.pt, then loads the published pretrained policy with the Kit
visualizer. It captures a screenshot and a 15-second X11 recording after the
playback readiness marker. The host needs xdotool, xprop, and ffmpeg. The runner
requires one newly visible Isaac Lab window and captures that window. Keep it
in the foreground throughout recording. Missing, ambiguous, hidden, or
covered windows fail the proof. Inspect the screenshot and a video frame to
confirm that Go2 and the terrain rendered.
Use a local image that includes Isaac Lab 3.0.0-EA.
The runner reuses it without rebuilding and leaves the Compose service running
so the logs and checkpoint remain available inside the container.
Its camera-following CLI overrides work in 3.0.0-EA but are deprecated. Use
KitVisualizerCfg in custom task code, as shown in docs/go2-ws/index.md.
python3 tests/workspace_smoke/go2_rough_smoke.py --workspace go2_ws
If only the fully built template image is available, pass
--workspace template_ws. Pass --build to build the chosen workspace first.
Report the summary JSON, training log, checkpoint, screenshot, and MP4 paths.
Signals
- GitHub stars
- 47
- Forks
- 10
- Last commit
- Oct 2026
Advanced
- Item type
- skill
- Key
go2-workspace-testing- Source
- github.com/j3soon/ros2-essentials
github.com/j3soon/ros2-essentials