القائمة
Use to fetch the contestant's best score and the per-board leaderboard for the Simulation Challenge. Read-only; safe to run without confirmation.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Download the Simulation Challenge LeRobot v2.1 training datasets from ModelScope using ./scripts/download_dataset.sh. Pulls task-suite data (instruction / manipulation / sim2real) from the agibot_world/GenieSim3.0-Dataset repo into a local dir. Trigger: When the user asks to "下载训练数据", "下载数据集", "拉取 lerobot 数据", "download training data", "download the dataset", "get the lerobot v2.1 data", "download task suite", mentions download_dataset.sh, ModelScope GenieSim3.0-Dataset, or needs training data to train a challenge model.
Launch a data_collection automated trajectory-collection task on a GPU host using the `geniesim autocollect run` CLI verb (which wraps scripts/run_data_collection.sh: docker run -d + in-container server+client). Trigger: when the user asks to "采集数据", "跑数据采集", "run data collection", "collect a task", "生产轨迹", "launch a tasks/geniesim_2025/<...>.json", or wants to produce agibot-format episodes from a data_collection task template.
Provision and launch the Simulation Challenge baseline inference model end to end: clone the inference code from a given git repo/branch, download the checkpoints from ModelScope into the repo's local checkpoints path, install deps, and start the inference agent. Trigger: When the user asks to "拉取 baseline 模型", "下载推理代码/权重", "搭一个 baseline", "clone the inference repo", "download ckpts", "set up the baseline model", "跑起来 baseline", "provision the baseline", or hands over a repo URL to stand up an inference server. Load even on light phrasings: "部署一下 demo", "部署 demo", "跑个 demo", "deploy a/the demo", "下载一下 baseline", "下载 baseline", "拉一下 baseline".
Entry point for the Simulation Challenge skill set. Use when the user mentions the challenge, leaderboard, submitting a model, or any of the /api/challenge/* endpoints — this skill picks the right downstream skill for them.
Reference for the Simulation Challenge inference wire protocol — the exact obs (input) and action (output) message format exchanged between the gateway/genie-sim simulator and the contestant's inference agent over the reverse WebSocket tunnel. Covers the JSON-RPC envelope, image encoding, per-board state/action joint layout, the plain-msgpack response caveat, and the authoritative source files. Trigger: When the user asks about "推理接口协议", "inference protocol", "obs/action format", "观测/动作格式", "网关下发什么", "agent 返回什么", "what does the gateway send", "result envelope", "state layout", "动作怎么拆", or needs to implement/debug the obs→model→action adapter in a tunnel agent.
Use when the contestant needs to obtain or refresh their Simulation Challenge JWT (CHALLENGE_TOKEN), or wants to inspect the current logged-in user. Trigger words include "log in", "token", "401", "current user", "name", "refresh".
| name | challenge-ranking |
| description | Use to fetch the contestant's best score and the per-board leaderboard for the Simulation Challenge. Read-only; safe to run without confirmation. |
All endpoints here are read-only.
Board scoping. Best-score, rank, and leaderboard are all per-board: there is no global "total" score, and the same user can have four independent best-score / rank tuples (one per board:
instruction/spatial/manip/robust). Always ask the user which board they care about, or report all four — never silently average.
CHALLENGE_TOKEN set for best-score (else → challenge-login)./leaderboard, /leaderboard/top5, /leaderboard/citation) are public — they work without a token, but include Authorization when you have one to stay symmetric with the rest of the skill set.Every Bash call should start with:
[ -f ~/.simubotix-challenge.env ] && . ~/.simubotix-challenge.env
best-score requires a board query parameter (it filters the user's per-board best):
[ -f ~/.simubotix-challenge.env ] && . ~/.simubotix-challenge.env
curl -fsS "$BASE_URL/api/challenge/best-score?board=instruction" \
-H "Authorization: Bearer $CHALLENGE_TOKEN" | jq
# { "status": "ok", "score": 573.0, "rank": 5 }
Replace instruction with spatial / manip / robust as needed. Report score and rank plainly. If status is "error" (or rank is 0), the user has no scored submission on that board yet — suggest they run challenge-submit-job against it.
To get a quick all-board summary, fan out in parallel:
for b in instruction spatial manip robust; do
printf '%-12s ' "$b"
curl -fsS "$BASE_URL/api/challenge/best-score?board=$b" \
-H "Authorization: Bearer $CHALLENGE_TOKEN" | jq -c '{score, rank}'
done
curl -fsS "$BASE_URL/api/challenge/leaderboard?board=instruction&page=1&per_page=20" \
-H "Authorization: Bearer $CHALLENGE_TOKEN" | jq
Each row is one user's best job for that board. Optional query params: sort (default score), order (asc / desc, default desc), q (substring search over user / organization name).
There is no rank_change field — only the current rank/score. Do not invent deltas.
For a homepage-style overview without four separate calls:
curl -fsS "$BASE_URL/api/challenge/leaderboard/top5" | jq
# { "boards": { "instruction": [...top 5], "spatial": [...], "manip": [...], "robust": [...] } }
Public, no auth needed.
curl -fsS "$BASE_URL/api/challenge/leaderboard/instruction/detail?user_uid=<uid>" | jq
The {board} path segment must be one of instruction / spatial / manip / robust. user_uid is the target user's UID (find it via the leaderboard row).
curl -fsS "$BASE_URL/api/challenge/leaderboard/citation" | jq -r '.citation // .'
Returns the platform-configured BibTeX string verbatim — surface as-is.
If the user wants their own row and they're past page 1, walk pages until you hit their name. Keep per_page ≤ 100 to stay polite.
challenge-submit-job (mind the 4-submissions-per-day quota; each board needs its own submission).challenge-poll-result.