executing-plans
executing-plans
Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available
experiment-launchersRoutes safe command-generation and launcher preset workflows for CUT, FastCUT, SinCUT, and pretrained experiments.
Independent GitHub snapshot · Automated classification · Original documentation governs
Use this sub-skill when the user asks about python -m experiments, launcher presets, or how to list the concrete training/testing command strings for the repository's known experiment families.
grumpifycat, pretrained, or singleimage.id argument means.dry path in this checkout.train.py or test.py commands.python -m experiments <name> <cmd> <id...>.grumpifycat, pretrained, and singleimage.run, train, test, run_test, print_names, and print_test_names.../translation-workflows/ for that.../data-preparation/ for that.placeholder_launcher.py as a supported workflow; it references stale or unsupported options.Run scripts/list_experiment_commands.py to inspect known presets safely:
python scripts/list_experiment_commands.py --family grumpifycat --kind train
python scripts/list_experiment_commands.py --family pretrained --kind test --ids 2,3
python scripts/list_experiment_commands.py --family singleimage --kind both --json
The helper prints command strings only; it never starts tmux, allocates GPUs, downloads datasets, or runs model training.
references/launcher-reference.md for preset command tables and CLI behavior.references/troubleshooting.md for the broken dry command, missing id, and tmux/GPUtil issues.../translation-workflows/references/cli-reference.md when you need to edit the commands the launchers emit.print_names and print_test_names were verified as safe inspection commands.dry command fails in this checkout because its method call omits required ids.launch, launch_test, stop, and close can create or kill tmux windows; do not use them merely to inspect command strings.