draw-thing

name	draw-thing
description	Generate local AI images and short media with Draw Things CLI on macOS. Use when you need local txt2img, img2img, model setup, imports, prompt refinement, or rig-aware best-model selection. NOT for UI implementation (design), ad copy iteration (ad-creative), or broad vendor/tool research (research).
argument-hint	<mode> [prompt or path] [--model qwen|z-image|flux2|video]
model	opus
license	MIT
metadata	{"author":"wyattowalsh","version":"1.1.0"}
allowed-tools	Bash Read Glob

Draw Thing

Local AI image and media generation through draw-things-cli. Use the current Draw Things command surface: generate, models list, models ensure, models import, and train lora.

Scope: Local Draw Things image generation, image editing, model setup, model inventory, imports, and short local media generation. NOT for UI implementation (design), ad creative strategy (ad-creative), cloud image APIs, or general AI-model market research.

---

Canonical Vocabulary

Term	Meaning	NOT
txt2img	Text-to-image generation: prompt in, image file out	img2img
img2img	Image-to-image generation: input image plus prompt	txt2img
media	Video or audio-video generation with --frames and video output	UI animation
model id	Draw Things model filename or model reference accepted by --model	display name guess
recommended pack	Current rig-aware model set approved for this Mac	every installed model
legacy model	Older model kept for compatibility, not recommended by default	deletion target
cfg	Current CLI guidance flag: --cfg	old --guidance-scale
strength	Img2img change amount: lower preserves, higher redraws	output scale
seed	Reproducibility value for identical prompts/settings	model selection
config file	JSON file passed with --config-file for advanced app settings	direct CLI flag

---

Dispatch

$ARGUMENTS	Mode	Action
generate <prompt> / create <prompt>	Generate	txt2img with best current image defaults
Natural language image description	Auto: Generate	Detect prompt intent and choose a model
edit <path> <prompt> / transform <path>	Edit	img2img with --image and --strength
Path to image + modification intent	Auto: Edit	Preserve original, write a new output file
video <prompt> / media <prompt> / animate <prompt>	Media	Generate short local media with --frames
setup / download models / best models	Setup	Inventory rig and ensure recommended pack
inventory / models / model status	Inventory	Run model inventory helper and summarize gaps
import <path-or-url>	Import	Use draw-things-cli models import
model <name>	Model info	Show current role, settings, and status
batch <prompt> / variations <prompt>	Batch	Use a shell loop with multiple seeds and outputs
refine / iterate	Refine	Re-run with a locked seed and one changed parameter
upscale <path> / controlnet ... / lora ...	Advanced	Verify config-file workflow before running
(empty)	Help	Verify CLI, show modes and examples

Auto-Detection Heuristic

1. Prompt asks to install, download, configure, or choose best models -> Setup.
2. Prompt asks for model list, current downloads, or missing models -> Inventory.
3. Prompt asks for video, motion, mp4, mov, frames, or animation -> Media.
4. Existing file path plus edit/restyle/change/transform -> Edit.
5. Existing file path plus upscale/control/lora/inpaint -> Advanced, then verify current CLI/config support before running.
6. Descriptive text with no file path -> Generate.
7. Ambiguous -> ask which mode and state the tradeoff.

---

Prerequisite Protocol

Run before any generation or model mutation:

1. Check CLI: command -v draw-things-cli.
2. If missing, show install command and stop:

   brew install drawthingsai/draw-things/draw-things-cli
   

3. Verify installed command surface: draw-things-cli generate --help.
4. For setup/inventory, prefer:

   uv run python skills/draw-thing/scripts/model_inventory.py --format json
   

5. Always pass --output; without it, the current CLI may only preview in supported terminals instead of writing a file.

---

Current Best Model Policy

As of 2026-05-15, this Mac is Apple M1 Pro, arm64, macOS 26.5, 32 GiB unified memory. Use current Draw Things optimized checkpoints and do not recommend legacy SDXL/SD1.5/FLUX.1 families unless the user explicitly asks for compatibility.

Role	Model id	Status	Default use
Fast image default	z_image_turbo_1.0_q8p.ckpt	recommended	Fast drafts, 8 steps, --cfg 0
Best text/layout/realism	qwen_image_2512_q8p.ckpt	recommended	Posters, typography, polished images
Creative base	z_image_1.0_q8p.ckpt	recommended	General creative exploration
High-end FLUX/reference/edit	flux_2_dev_q6p.ckpt	recommended	FLUX.2 workflows on 32 GiB
Instruction image edit	qwen_image_edit_2511_q8p.ckpt	recommended	Img2img/edit prompts
Art/prompt-following alternative	hidream_i1_full_q5p.ckpt	recommended download	Highest-quality art alternative
Fast local media	ltx_2.3_22b_distilled_q6p.ckpt	recommended download	Short text/video generation
Rig-fit Wan media	wan_v2.2_5b_ti2v_q8p.ckpt	recommended download	Current Wan video on 32 GiB

Legacy/redundant downloaded models are not deletion targets without approval: flux_1_schnell_q8p.ckpt, flux_1_fill_dev_q8p.ckpt, and flux_2_klein_4b_q6p.ckpt.

Auxiliary text/VLM files and incomplete partials are not recommended generation models. For example, llama_3.1_8b_instruct_q8p.ckpt.part is an incomplete stale auxiliary artifact, not a mid-May 2026 best-pack model.

Qwen Layered remains a current optional workflow target, but not a guaranteed best-pack install on this rig: qwen_image_layered_1.0_bf16_q6p.ckpt failed verification with a checksum mismatch, and qwen_image_layered_1.0_bf16_q8p.ckpt repeatedly stalled during CLI download. Use qwen_image_edit_2511_q8p.ckpt for reliable local editing unless the user explicitly asks to keep retrying Layered.

For exact install commands, current/legacy distinctions, and quantization rules, load references/model-catalog.md.

---

Core Generation Protocols

Every mode follows this pattern:

1. Validate: CLI exists, input files exist, model is downloaded or --download-missing is intentional.
2. Select defaults: choose a current model by role unless the user provided --model.
3. Build command: use only flags shown by draw-things-cli generate --help, or use --config-file for advanced app settings.
4. Show command: display the full command before running.
5. Execute: run through Bash and capture output.
6. Report: output path, seed, model id, dimensions, steps, and cfg.

Mode: Generate

Fast draft default:

draw-things-cli generate \
  --model z_image_turbo_1.0_q8p.ckpt \
  --prompt "<prompt>" \
  --steps 8 \
  --cfg 0 \
  --width 1024 \
  --height 1024 \
  --seed <seed> \
  --output <output.png>

Quality/text default:

draw-things-cli generate \
  --model qwen_image_2512_q8p.ckpt \
  --prompt "<prompt>" \
  --steps 40 \
  --cfg 4 \
  --width 1328 \
  --height 1328 \
  --seed <seed> \
  --output <output.png>

Mode: Edit

draw-things-cli generate \
  --model qwen_image_edit_2511_q8p.ckpt \
  --image <input.png> \
  --prompt "<edit instruction>" \
  --strength 0.55 \
  --steps 30 \
  --cfg 4 \
  --seed <seed> \
  --output <output.png>

Use --strength 0.25-0.4 for subtle edits, 0.5-0.65 for visible edits, and 0.75+ for heavy redraws.

Mode: Media

draw-things-cli generate \
  --model ltx_2.3_22b_distilled_q6p.ckpt \
  --prompt "<motion prompt>" \
  --frames 49 \
  --width 768 \
  --height 512 \
  --seed <seed> \
  --output <output.mov>

Use LTX distilled first for this rig. Keep width/height divisible by 32 and frames divisible by 8 plus 1.

Mode: Setup

Run inventory first:

uv run python skills/draw-thing/scripts/model_inventory.py --recommended-pack all --format json

Then run missing ensure_command entries after showing them to the user. Do not delete legacy models unless the user separately approves a prune operation.

Mode: Inventory

uv run python skills/draw-thing/scripts/model_inventory.py --format json

Summarize installed, missing recommended models, and redundant legacy downloads. Report incomplete_downloads and auxiliary_or_dependency_files separately so helper files or stale .part files are not mistaken for current best models.

Mode: Batch

The current CLI does not expose --batch-count. Use a shell loop and explicit output names:

for seed in 1201 1202 1203 1204; do
  draw-things-cli generate \
    --model z_image_turbo_1.0_q8p.ckpt \
    --prompt "<prompt>" \
    --steps 8 \
    --cfg 0 \
    --width 1024 \
    --height 1024 \
    --seed "$seed" \
    --output "$HOME/Pictures/draw-thing/run-$seed.png"
done

Mode: Advanced

The current generate --help does not expose old direct flags such as --controls, --loras, --mask, --upscaler, --batch-count, --sampler, or --guidance-scale. For ControlNet, LoRA, inpaint, upscale, tiling, and other app-level settings:

1. Re-check draw-things-cli generate --help.
2. Use Draw Things app-managed settings or --config-json / --config-file only when the schema is verified.
3. Never invent direct flags from older examples.

---

Prompt Quick-Reference

Model family	Style	Negative prompt
Z Image Turbo/Base	Direct natural-language instructions; concise composition details	Supported when useful, keep short
Qwen Image 2512	Explicit layout, typography, exact text, editorial language	Supported, keep targeted
Qwen Edit/Layered	Imperative edit instructions and preservation constraints	Use only for what to avoid changing
FLUX.2	Natural language, reference/edit details, precise scene intent	Verify per model; do not assume FLUX.1 rules
HiDream I1 Full	Rich art direction, style, medium, lighting, composition	Short targeted negatives if used
LTX/Wan media	Describe motion over time, camera movement, temporal continuity	Avoid still-image-only tag lists

For templates and troubleshooting, load references/prompt-patterns.md.

---

Scaling Strategy

Request size	Strategy
Single image	Use z_image_turbo_1.0_q8p.ckpt unless quality/text needs Qwen 2512
Text/layout critical	Use qwen_image_2512_q8p.ckpt and explicit text/layout prompt
Edit existing image	Use qwen_image_edit_2511_q8p.ckpt with conservative strength first
Layer/editability workflow	Use Qwen Layered only if already installed; otherwise explain prior install failures and ask before retrying
Heavy art quality	Try hidream_i1_full_q5p.ckpt; fall back if memory pressure appears
Short video	Use LTX distilled q6p first; keep dimensions moderate
Many variations	Use shell loops, fixed output paths, and modest seed batches

---

Output Handling

• Default skill output directory: ~/Pictures/draw-thing/.
• Create it before generation: mkdir -p "$HOME/Pictures/draw-thing".
• Always pass --output <file>; do not rely on implicit output behavior.
• Never overwrite source images; use descriptive new filenames.
• Always report output file, seed, model, steps, cfg, dimensions, and whether the model was downloaded during the run.

---

Error Recovery

Error	Likely cause	Action
unknown flag	Old guidance or CLI mismatch	Run draw-things-cli generate --help; remove stale direct flag
model not found	Wrong id or not downloaded	Run draw-things-cli models list; use models ensure --model <id>
no output file	Missing --output or preview-only terminal mode	Re-run with explicit file path
killed/OOM	Model too large or dimensions too high	Lower dimensions, use q6p/q5p, or switch to faster model
poor text	Wrong model/prompt for typography	Use Qwen Image 2512 with exact quoted text
video fails	Model missing or memory pressure	Ensure LTX q6p, reduce dimensions/frames

---

Reference Files

Load ONE reference at a time.

File	Content	Load When
references/cli-reference.md	Current command surface, flags, models subcommands, config caveats	Building non-trivial commands
references/model-catalog.md	2026-05-15 current best model pack, exact ensure commands, legacy notes	Choosing/downloading models
references/prompt-patterns.md	Prompt patterns for Qwen, Z Image, FLUX.2, HiDream, LTX/Wan	Complex prompts or quality issues
references/controlnet-guide.md	Advanced controls/import/config guidance with stale-flag warnings	ControlNet, LoRA, inpaint, upscale
references/workflow-recipes.md	Current recipes using exposed flags and shell loops	Multi-step creative workflows

---

Validation Contract

Run from this skill directory before declaring changes complete:

python scripts/check.py
python scripts/model_inventory.py --format json

Completion criteria:

1. scripts/check.py exits 0.
2. Model inventory identifies recommended-pack status.
3. No command examples use stale direct flags unsupported by current generate --help.
4. No portable-CLI violations remain under this skill directory.

---

Critical Rules

1. Always check command -v draw-things-cli before generation or model mutation.
2. Always use draw-things-cli generate --help as the source of truth for direct flags.
3. Always pass --output; preview-only output is not enough for agent workflows.
4. Use --cfg, not stale --guidance-scale.
5. Do not use stale direct flags: --sampler, --batch-count, --upscaler, --mask, --controls, or --loras unless current help/config schema proves support.
6. Prefer the recommended 2026 rig pack over legacy SDXL, SD1.5, SD2.x, SVD, FLUX.1, Qwen 1.0, Qwen Edit 2509, Wan 2.1, or old community checkpoints.
7. Never delete or prune models without separate explicit user approval.
8. Show the full command before running it.
9. Preserve source images; write new output files.
10. Report seed, model id, output path, dimensions, steps, and cfg.
11. For video, keep dimensions moderate and frame count divisible by 8 plus 1.
12. For advanced ControlNet/LoRA/inpaint/upscale workflows, verify --config-json or --config-file schema before execution.
13. Do not treat auxiliary text/VLM files, sidecars, or incomplete .part downloads as recommended generation/media models.