UI はローカライズ済みです。スキルの紹介と説明文書は選択した言語へ AI が自動翻訳し、コード・コマンド・パス・専門用語は原文のままです。原作者の文書が優先されます。
スキルを探す
生産性・共同作業 · SKILL.md

pipeline-cli

pipeline-cli

Guides Raster Vision agents through rastervision CLI pipeline runs, pipeline configs, runners, model bundles, and config or URI troubleshooting.

Independent GitHub snapshot · Automated classification · Original documentation governs

スキルの説明

GitHub

name: pipeline-cli description: "Guides Raster Vision agents through rastervision CLI pipeline runs, pipeline configs, runners, model bundles, and config or URI troubleshooting." disable-model-invocation: true metadata: disco-role: operating license: NOASSERTION

pipeline-cli

Use this sub-skill when a Raster Vision request is about:

  • rastervision run, run_command, predict, or predict_scene
  • PipelineConfig / RVPipelineConfig loading, output URIs, or get_config(s) modules
  • --arg, --splits, runner selection, or config summary checks
  • model bundles, prediction inputs, or CLI failures caused by config or URI mismatches

Read first

  1. CLI reference
  2. Pipeline configuration
  3. Model bundles
  4. Troubleshooting

Runtime helper

Use scripts/summarize_pipeline_config.py to inspect a Raster Vision config module or JSON file without executing any pipeline commands.

Boundaries

  • Keep this sub-skill to Raster Vision CLI and generic pipeline execution.
  • Route core data, source, and label APIs to the data-and-models sub-skill.
  • Route PyTorch task recipes to the pytorch-workflows sub-skill.
  • Route AWS, Docker, and CloudFormation details to the cloud-and-filesystems sub-skill.
  • Do not include runner internals beyond CLI-level routing.
保存したスキル