메뉴
Use when working on NPA reference SkyPilot YAMLs, runner scripts, cookbooks, or customer-adaptable pipeline implementations.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Use when ingesting, validating, curating, or querying production sensor data as a versioned dataset-of-record, or wiring the dataset-ingest-curate workflow.
Use when generating adversarial scenarios via RL, ranking mined failures of a policy-under-test, or wiring the adversarial-scenario-hardening workflow.
Use when navigating the Cosmos3 integration in NPA or locating upstream Cosmos3 framework files, defaults, scripts, configs, recipes, and docs.
Use when Cosmos3 setup, fetch, inference, CUDA, uv, Docker, Hugging Face, GitHub, NGC, or checkpoint staging fails in NPA or in an upstream Cosmos framework checkout.
Use when setting up Cosmos3 access through NPA, checking source or Hugging Face reachability, staging the public Cosmos3 framework and checkpoint cache, or deciding which NPA workflow to use before inference.
Use for Claude Code reviews that need robotics, simulation, GPU-routing, sim-to-real, or BDD100K pipeline domain context.
| name | workbench-reference-workflows |
| description | Use when working on NPA reference SkyPilot YAMLs, runner scripts, cookbooks, or customer-adaptable pipeline implementations. |
The supported, customer-facing catalog is the
npa.workflowspec set undernpa/workflows/workbench/npa-workflows/. The raw SkyPilot task YAMLs below are internal runtime templates relocated tonpa/src/npa/workflows/skypilot/; they back therun_*.pywrappers and SkyPilot-only capabilities, and must not be re-added to the shownnpa/workflows/workbench/catalog (guardrail-enforced).
Use this skill for repository workflow YAMLs, runner scripts, cookbooks, artifact contracts, and customer-adaptable pipeline implementations.
npa/src/npa/workflows/skypilot/.bdd100k-pipeline.yaml: BDD100K ingest, backfill, CLIP embedding,
materialized views, training, and evaluation.cosmos3-ea-fetch.yaml: Cosmos3 source/checkpoint fetch.cosmos3-text-to-image-inference.yaml: H100 text-to-image smoke inference.isaac-lab-rl-train.yaml: single Isaac Lab RL job.isaac-lab-rl-sweep.yaml: Isaac Lab parameter sweep.mjlab-eval.yaml: MJLab evaluation.retargeting.yaml: motion retargeting.sim-to-real-loop.yaml: iterative sim-to-real loop.sim-to-real-pipeline.yaml: full sim-to-real pipeline.sim-to-real-trigger.yaml: trigger wrapper for sim-to-real work.sonic-train-standalone.yaml: standalone SONIC training.sonic-export.yaml: SONIC export.sonic-eval.yaml: SONIC evaluation.sonic-export-eval.yaml: export plus evaluation.sonic-locomotion-finetuning.yaml: retargeting, SONIC, and MJLab flow.vlm-eval.yaml and vlm-eval-benchmark.yaml: VLM evaluation loops.npa workbench workflow ... and tool-specific workflow commands
such as npa workbench mjlab workflow or npa workbench retargeting workflow.envs does not support self-referencing interpolation. Use explicit
values and comments for alternatives.sky jobs launch has no dry-run flag in the pinned path. Use local YAML
parsing, command help, and mock-endpoint tests before live submission.npa/.venv/bin/python -m pytest npa/tests/guardrails/test_skills_index.py -q
The smoke test parses the listed workflow YAMLs and invokes workflow CLI help.