---

name: deployment-and-interfaces description: "Routes min(DALL·E) command-line, notebook, Tkinter GUI, and Replicate/Cog interface workflows beyond direct MinDalle API calls." disable-model-invocation: true metadata: disco-role: operating license: MIT

deployment-and-interfaces

Use this sub-skill when the task is about min(DALL·E) public interfaces outside a direct MinDalle API call: command-line usage, Colab/notebook patterns, the Tkinter UI, Replicate/Cog deployment, predictor filenames, output paths, or interface-specific troubleshooting.

Start here

• Read references/cli-reference.md for the original image_from_text.py flags, defaults, save behavior, ASCII preview, and the verified absence of an installed console command.
• Read references/interface-workflows.md for the Colab/notebook flow, Tkinter controls and limits, Replicate predictor inputs/defaults/ranges, filename sanitation, output extensions, Cog pins, and deployment caveats.
• Read references/troubleshooting.md when the user reports a missing command, slow downloads/inference, CPU/FP16 problems, display errors, Cog/CUDA issues, unexpected Replicate filenames, or progressive-output counts.

Bundled helpers

• scripts/min_dalle_cli_template.py is the safe replacement for the original command-line script. It defaults to --dry-run; add --run only when the user explicitly wants model construction, possible downloads, and inference.
• scripts/replicate_filename_sanitize.py previews the Replicate-style output basename without requiring Cog or model loading.

Useful no-network checks:

python scripts/min_dalle_cli_template.py --help
python scripts/min_dalle_cli_template.py --text "artificial intelligence" --no-mega --top-k 256
python scripts/replicate_filename_sanitize.py --self-test
python scripts/replicate_filename_sanitize.py --text "Dali painting of WALL·E"

Boundary routes

• For direct MinDalle construction, generate_image, generate_image_stream, sampling arguments, tensor/PIL image semantics, and reusable model lifetime, route to ../text-to-image-generation/SKILL.md.
• For model weight downloads, models_root, cache layout, device selection, dtype/backend behavior, CUDA/CPU trade-offs, and environment setup, route to ../model-assets-and-runtime/SKILL.md.
• Do not tell future agents to run the original repository scripts or notebook. Use this sub-skill's bundled helpers and distilled references instead.