agentleFS
Sign inStart free
mehbylabsmehbylabs/creator-skills/skills/creator-image/SKILL.md

creator-image

Skill7 starsChanged 16 days agoMITView on GitHub

Generate or edit images with whatever the user has — Higgsfield CLI, Google Gemini API (Nano Banana), OpenAI API (gpt-image), Antigravity CLI (agy), Codex CLI, or the Gemini web app / Google Flow via computer use — with automatic fallback. Use when the user asks to generate, create, draw, or edit an image, illustration, hero image, banner art, poster, wallpaper, icon art, sticker, transparent PNG, or to restyle or edit a photo with a reference. Not for thumbnails with headline text (creator-thumbnail), product shots (creator-product-photo), consistent characters (creator-character), video, or setup.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If the extension changed (e.g.

    line 53
  2. Distinctive:

    Never ask for API keys in chat.

    line 64

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Creator Image

Step 1 — Shape the request

Decide from the message; ask at most one question if something essential is missing.

  • Prompt — rewrite with references/image-prompting.md (medium, subject, composition, light, palette, constraints). Keep the user's intent.
  • Ratio — banner/hero/YouTube 16:9; Story/Reel/Short/phone 9:16; Instagram feed 4:5; Pinterest 2:3; otherwise 1:1.
  • Size — 1k default; 2k/4k only when asked or for print/hero.
  • References — attached or named images → --ref (repeatable).
  • Transparency — logos, stickers, cutouts → --transparent.
  • Output — the user's path, else ./out/<slug>.png.
  • Provider — --provider only when the user names one ("use agy", "with codex").

Step 2 — Generate

python3 scripts/creator.py image --prompt "<prompt>" --out <path> --ratio <r> [--size 2k] [--ref a.png] [--transparent] [--provider <id>]

Agent-CLI routes take 30–120 s; run in the background if your harness supports it.

Step 3 — Act on the JSON result

statusDo
okView the image and check it against the brief. Report path + provider in one line
ok + "post": "remove_background"No provider could do alpha; run python3 scripts/media.py cutout --image <path> --out <path> and say so
needs_browserFollow references/browser-<next[0]>.md using request; finish with creator.py adopt
blockedContent refused. Don't retry elsewhere; explain and offer a revised prompt
failedOne line per attempt (provider → reason) and the single most useful fix from connect — a command the user runs in their own terminal (references/connect.md)

If the extension changed (e.g. .png → .jpg without ffmpeg), mention it.

Step 4 — Iterate

Change one thing per round. Pin the provider that made the accepted image (--provider <id>) so the look stays stable. To edit the last output, pass it as --ref and describe only the change.

Rules

  • Be brief: no JSON, no command narration, no model ids unless asked.
  • Never ask for API keys in chat.
  • Text inside images: only short, quoted text, and only on providers good at it; otherwise generate clean art and add type with media.py text.

References

  • references/image-prompting.md — prompt structure, text, references
  • references/providers.md — matrix, order, fallback, exit codes
  • references/media-ops.md — cutout, text overlay, reframe
  • references/browser-gemini-browser.md, references/browser-flow-browser.md — manual routes

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free