メニュー
Use this skill when the user wants to run the end-to-end all-sky photo pipeline from a raw all-sky image to a replanned observing sequence, using the packaged run_pipeline.py entrypoint in this shared directory.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/Yu-Yang-Li/StarWhisper --skill photo-to-replanコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Compare the published sparse ZTF/ATLAS early-classification benchmark and lint a light-curve table against the varlen 3-30 / 7-class contract. Use for varlen, sparse light curves, XGBoost vs Transformer vs LSTM, seven merged variability classes, or Early Classification from Sparse Light Curves. Does not train; never treats a test-set score as an explosion time.
Evaluate an observing-policy comparison against the StarWhisper-Explore pre-registered bar and return pass/fail per criterion. Use for 决策边界, survey completeness vs transient follow-up, policy A/B comparison, stable negative result, GOAI Explore, or judging whether an agent policy actually cleared its threshold. Reads a metrics table; never simulates nights.
Validate an NGSS observing configuration, compute the night exposure budget, and lint a target list before anyone touches hardware. Use for observe_config.json, StarWhisper Telescope, NGSS, NINA, 夜计划, moon distance constraints, exposure count/time/wait, how many targets fit in a night, or the NGSS endpoint contract. Never sends a command.
SOC 職業分類に基づく
SKILL.md を表示中
| name | photo-to-replan |
| description | Use this skill when the user wants to run the end-to-end all-sky photo pipeline from a raw all-sky image to a replanned observing sequence, using the packaged run_pipeline.py entrypoint in this shared directory. |
Use this skill when the task is to go from a raw all-sky image to replanned observing outputs.
Locate the shared photo_to_replan_pipeline directory.
Run the single-entry pipeline script from that directory:
python run_pipeline.py --image /abs/path/to/image.jpg
Optional:
--position-name position_X--save-overlayoutput/<image_stem>/pipeline_report.jsonscenario.jsonreplanned_schedule.jsondeferred_targets.jsonreplanned_sequence.ninaTargetSetsky_replan_plot.pngYYYY_MM_DD_HH_MM_SS.jpg form.Pic2mask/full_image_infer/infer_full_image.py.sequence_adjust_tool/scheme_horizon_to_image/replan.py.configs/default.json.A run counts as pipeline-successful if these files exist and are readable:
pipeline_report.jsonscenario.jsonreplanned_schedule.jsonreplanned_sequence.ninaTargetSetBusiness success is separate. Check deferred_targets.json and replanned_schedule.json to see whether future slots were actually filled.