Menu

Individual skill

image-lab-2

0upvotes

Generate image alternatives with the price shown before any spend. Quick: one finished prompt becomes several alternatives across the 2-3 best-fit image models. Variation: controlled variations (camera, lighting, style, aspect) of a brief, optionally from a reference photo, in an Excel workbook where the user ticks the rows to generate. Creative Direction: several different directions for an open brief, one image each first. Each paid batch shows its total and needs one approval. Triggers include "generate that with Image Lab 2", "8 options of this with Image Lab 2", "run Image Lab 2 on this photo", "with Image Lab 2, try different lighting and camera angles", "with Image Lab 2, explore four creative directions". It has to be named with its version: a request that only says "Image Lab" for a single prompt belongs to the older image-lab skill, and a bare "generate an image" to native image generation.

examples/community/image-lab-2/SKILL.md

Read pinned source

Source commit d9efb25f4a1f7941edfd42961769e0836fb0565e

What was checked

Matches the observed repository HEAD
Yes
Scanner rules are current
No
Repository scan complete
No
Skill scan complete
No
Observed behavior
Not checked yet
Last checked
Nov 5, 2026

These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.

Install this skill

Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.

An install command is unavailable: current rules, complete repository and skill scans, an eligible behavior level and a permitted license are required.

Repository context

Repository stars (not skill ratings)
16
Repository last updated
Nov 5, 2026
License
Apache-2.0

Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.

Files and excerpt

File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.

Browse files
Read excerpt

# Image Lab 2

**See the price. Approve each paid batch. Generate in parallel. Keep a record of everything.**

Three modes share one engine (`scripts/image.py`), one ledger and one gate. **Quick**: one finished prompt, spread across the best-fit models, no plan. **Variation** and
**Creative Direction** (together "experiment mode"): a `plan.json` becomes `experiment.xlsx`, where the user ticks the rows to generate.

Setup: run commands from this folder (the one holding this file; for an installed skill, for example `~/.claude/skills/image-lab-2/`) (`python scripts/image.py <command>`); `pip install -r requirements.txt` (experiment mode; quick mode needs nothing); a gateway key in
`LOOMLOOM_TOKEN_COGFOUNDRY` (https://console.cogfoundry.ai/api-keys), needed only for `run`; the loomloom CLI is **not** used. On Windows set `PYTHONUTF8=1` (bash:
`PYTHONUTF8=1 python ...`; PowerShell: `$env:PYTHONUTF8=1`). Keep plans and experiments in a folder of your own (`./image-lab-out/`), never inside the skill folder: an update replaces it. Read `./out/<name>` below as that folder. If your packages are in a virtual environment, call its interpreter (`.venv/Scripts/python` on Windows) in place of `python`, with absolute paths for `--plan` and `--out`. Without `AskUserQuestion` or `SendUserFile`, ask in plain chat and give file paths.

## Start here

First time? Read [`references/quickstart.md`](references/quickstart.md) (one page): which route to take, the five-command loop, the plan file in 30 seconds, what the
Back to top