gpt-image-2-apilisted

Generate or edit images with XApex asynchronous tasks as the default route, AtlasCloud as the first edit fallback, and aifast as the final fallback, plus forced-provider profiles and batch generation. Use for text-to-image, local or URL reference editing, multi-reference composition, precision-sensitive graphics, XApex image-group tokens, provider-specific sizes and quality, resilient routing, and batch jobs.
kiakun-collab/kiakun-skills · ★ 1 · AI & Automation · score 57

Install: claude install-skill kiakun-collab/kiakun-skills

# GPT Image 2 API The scripts are already built and tested. **Run them directly — never rewrite, regenerate, or reimplement them.** Just call `node scripts/<name>.js` from this skill directory with the right flags. Reach for `references/api-reference.md` only when you need provider payloads or to diagnose an API error. ## Start Here (no setup needed on each run) ```bash # Text to image node scripts/generate.js --prompt "a red fox in snow" # Edit one image node scripts/edit.js --image photo.png --prompt "add sunglasses" # Batch: many prompts at once node scripts/batch.js --promptlist prompts.txt # Default: XApex asynchronous submission and polling node scripts/generate.js --prompt "a red fox in snow" --quality low node scripts/edit.js --image photo.png --prompt "add sunglasses" # Force synchronous XApex or a specific fallback provider only when needed node scripts/generate.js --sync --prompt "a lighthouse at dusk" node scripts/generate.js --profile vip --prompt "force aifast max" ``` Add `--dry-run` for a no-cost route/cost preview. Add `--json` for a machine-readable result. If a run fails with a missing-key or config error, do the one-time setup below. Otherwise skip it. ## Choosing a command | You want to... | Use | |---|---| | One image from text | `generate.js` | | One image edited from a reference | `edit.js` | | Many images in one go | `batch.js` | ## Reference-image intent (decide before editing) When the user supplies a reference image, decide **why** the