gooseworks-ai/goose-skills / skills/video/create-image-fal
create-image-fal
VerifiedGenerate or edit an image via any FAL image model (nano-banana edit, gpt-image, flux, ...), routed through the GooseWorks proxy, so the call is charged to your GooseWorks credits. image_urls must be public URLs; --image puts a local file on the fal CDN first. The recipe names the model + prompt. Use for keyframes, flat-cover transforms, product hero edits. (For the OpenAI gpt-image family specifically, create-image-gpt-image-fal also exists.)
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 8.0 KB |
| references/avatar-generation.md | markdown | 9.0 KB |
| scripts/brand_pack_lint.py | script | 4.4 KB |
| scripts/gen_image.py | script | 6.6 KB |
| scripts/image_quality.py | script | 1.5 KB |
| scripts/media_proxy.py | script | 56.8 KB |
| skill.meta.json | config | 177 B |
| tests/smoke-test.md | markdown | 280 B |
| tests/test_brand_pack_lint.py | script | 4.8 KB |
| tests/test_image_quality.py | script | 1.2 KB |
scripts/brand_pack_lint.py
View on GitHub#!/usr/bin/env python3
"""Refuse an image request that would show the brand's own product as a blank pack.
A 15 Oct test video asked GPT Image 2 for a food pouch with "its unprinted back facing the
camera" and "no printed words", and sent no product photo. The pack in the hands read as a
different product. The rule "no brand text in a prompt" bans TYPING label text; it never means
removing the product's real label.
lint_brand_pack(prompt, use, model, image_urls) -> list of problems ([] when it passes)
`use` says where the brand's product is in the picture:
- "static": on a surface or flat. The words below are refused; the real photo is usually
composited on afterwards (overlay).
- "held": held, opened, poured or used on screen. The words below are refused, the model
must be openai/gpt-image-2/edit, and the real product photo must be image 1, its front
facing the camera.
python brand_pack_lint.py --use held --model openai/gpt-image-2/edit \
--image https://.../pack-front.jpg --prompt "..." exit 1 with the problems
"""
import argparse
import json
import re
import sys
HELD_MODEL = "openai/gpt-image-2/edit"
PACK = (r"(?:pack|packet|package|packaging|pouch|box|bag|bottle|jar|tube|tub|carton|sachet|can|tin|"
r"wrapper|sleeve|label|front|side|back|lid|cap)s?")
# Words that ask for the advertised product without its real print.
BLANK_WORDS = [
(re.compile(r"\bun-?printed\b", re.I), "unprinted"),
(re.compile(r"\bun-?label(?:l)?ed\b|\blabel-?less\b|\blabel-?free\b", re.I), "unlabeled"),
(re.compile(r"\bblank\b(?:\W+\w+){0,3}?\W+" + PACK + r"\b|\b" + PACK + r"\W+(?:\w+\W+){0,2}?blank\b", re.I), "blank"),
(re.compile(r"\bno (?:printed )?(?:words|text|lettering|writing|print|label|logo|branding)\b(?:\W+\w+){0,2}?\W+(?:on|of)\b"
r"|\bno printed (?:words|text|lettering|writing)\b|\bwithout (?:a |any |its )?(?:label|print|printing|branding)\b", re.I),
"no printed words"),
(re.compile(r"\bback (?:of the \w+ )?(?:is |)(?:facing|faces|turned to(?:wards)?) (?:the )?(?:camera|lens|viewer)\b"
r"|\b(?:its|the) back (?:to|towards) the camera\b", re.I), "back facing the camera"),
]
FRONT = "Image 1 is the real product photo: keep its exact shape, colours and printed label, with its front facing the camera."
def blank_words(prompt):
"""The blank-pack words a prompt uses: [(label, matched text)]."""
found = []
for rx, label in BLANK_WORDS:
m = rx.search(prompt or "")
if m:
found.append((label, m.group(0)))
return found
def lint_brand_pack(prompt, use, model, image_urls, product_image=None):
"""Problems with an image request whose object is the brand's own product ([] when it passes)."""
if use not in ("static", "held"):
return ["--brand-product is static or held, not %r" % use]
problems = ["the prompt asks for the brand's product %s (%r). Show its real label from the real photo; "
"never ask for an unprinted, blank, unlabeled or back-facing pack. If the label garbles, reframe "
"so the hands cover it, or cut to the real photo." % (label, text) for label, text in blank_words(prompt)]
if use == "held":
if model != HELD_MODEL:
problems.append("a pack held or used on screen is made with %s (the real photo as image 1), not %s"
% (HELD_MODEL, model))
if not product_image:
problems.append("a pack held or used on screen needs --product-image, the real product photo, sent as image 1")
elif image_urls and image_urls[0] != product_image:
problems.append("the real product photo must be image 1; it is not first in image_urls")
return problems
def with_front_line(prompt):
"""The prompt with the front-facing line added once."""
return prompt if FRONT in prompt else (prompt.rstrip() + "\n\n" + FRONT)
def main():
ap = argparse.ArgumentParser(description=__doc__.split("\n")[0])
ap.add_argument("--prompt", required=True)
ap.add_argument("--use", required=True, choices=["static", "held"])
ap.add_argument("--model", required=True)
ap.add_argument("--image", action="append", default=[], help="image_urls in order")
a = ap.parse_args()
problems = lint_brand_pack(a.prompt, a.use, a.model, a.image, a.image[0] if a.image else None)
print(json.dumps({"pass": not problems, "problems": problems}, indent=1))
sys.exit(1 if problems else 0)
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse 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 | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |