How do I install the Generating Character And Background Images skill?

Run npx skillmds@latest add divakarwl12/generating-character-and-background-images in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Generating Character And Background Images safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Generating Character And Background Images?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Generating Character And Background Images free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Generating Character And Background Images?

divakarwl12 (@divakarwl12) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Generating Character And Background Images by divakarwl12

Name: Generating Character And Background Images
Author: divakarwl12

Generating Character And Background Images

Reads characters.json and backgrounds.json, then generates reference images for each character and background using the kie.ai nano-banana-pro API. Character images (1:1 aspect ratio) are saved to ./characters/ and background images (16:9) to ./backgrounds/, named by their IDs (char_001.png, bg_001.png, etc.). Each image is then uploaded to imgbb for permanent public hosting, and the imgbb URL is written back as an image_url field in both JSON files — these URLs are required by generate_videos.py. Part of the Story-to-Animation pipeline (Step 3 of 5). Run scripts/generate_images.py to execute. Use when both JSON prompt files exist and the user wants to generate the actual reference images. After generating all images, ALWAYS present the output and wait for explicit user approval. Do NOT automatically trigger the next pipeline step.

divakarwl12 Updated Sep 17, 2026

File contents

Character & Background Image Generation

Generate reference images with kie.ai nano-banana-pro, host them on imgbb, and write permanent image_url fields back into the JSON files.

Setup: .env

On first run the script automatically creates a .env template in your project directory and exits. Open it, fill in your keys, then re-run:

KIE_API_TOKEN=your_kie_api_key_here
IMGBB_API_KEY=your_imgbb_api_key_here

KIE_API_TOKEN — from kie.ai dashboard → API Keys
IMGBB_API_KEY — from api.imgbb.com (free account)
Add .env to .gitignore — never commit API keys
Real environment variables always take precedence over .env values

A reference copy is at assets/.env.example in this skill's folder.

Prerequisites

characters.json and backgrounds.json must exist (from Skill 2)
.env in project directory with both keys filled in (auto-created on first run)
pip install requests

Running the Script

cd /path/to/your/project
python ~/.claude/skills/generating-character-and-background-images/scripts/generate_images.py

What the script does per image:

Generate — POST to jobs/createTask (model: nano-banana-pro, aspect_ratio, resolution: 1K)
Poll — GET jobs/recordInfo?taskId=... until state: success; extract resultUrls[0]
Download — save PNG to ./characters/{id}.png or ./backgrounds/{id}.png
Upload to imgbb — POST base64-encoded PNG to api.imgbb.com/1/upload; get permanent URL
Write back — store imgbb URL as image_url in characters.json / backgrounds.json

Character images: 1:1 aspect ratio (reference sheets) Background images: 16:9 aspect ratio (cinematic wide shots)

If imgbb upload fails, the kie.ai URL is used as fallback (may not be permanent).

Regenerating Specific Images

Update the prompt field in the JSON
Delete the existing PNG (e.g., ./characters/char_001.png)
Re-run the script — it skips existing files, regenerates deleted ones

Important: image_url Fields

After running, each JSON entry has an image_url field containing a permanent imgbb URL. Skill 5 (generate_videos.py) reads these URLs to generate composite images — do not delete or overwrite them.

Review Gate (MANDATORY)

After all images are generated, present this EXACTLY:

✅ Image Generation complete.

📋 Summary:
- Character images: [X] generated → ./characters/
  [char_001.png: "Name", char_002.png: "Name", ...]
- Background images: [X] generated → ./backgrounds/
  [bg_001.png: "Name", bg_002.png: "Name", ...]
- Failed: [X] (list any failures)
- image_url (imgbb) written to: characters.json, backgrounds.json

👉 Please review the generated images. You can:
  - Approve all → say "approved" or "proceed"
  - Reject specific images → e.g., "regenerate char_001, the hair is wrong"
    (update prompt in characters.json, delete the PNG, re-run script)
  - Replace images manually → save your PNG as ./characters/char_001.png,
    then tell me to upload it to imgbb to get a fresh image_url

⏸️ Waiting for your approval before creating the shot list.

NEVER proceed to the next skill automatically. Wait for explicit approval.

---name: generating-character-and-background-imagesdescription: Reads characters.json and backgrounds.json, then generates reference images for each character and background using the kie.ai nano-banana-pro API. Character images (1:1 aspect ratio) are saved to ./characters/ and background images (16:9) to ./backgrounds/, named by their IDs (char_001.png, bg_001.png, etc.). Each image is then uploaded to imgbb for permanent public hosting, and the imgbb URL is written back as an image_url field in both JSON files — these URLs are required by generate_videos.py. Part of the Story-to-Animation pipeline (Step 3 of 5). Run scripts/generate_images.py to execute. Use when both JSON prompt files exist and the user wants to generate the actual reference images. After generating all images, ALWAYS present the output and wait for explicit user approval. Do NOT automatically trigger the next pipeline step.---# Character & Background Image GenerationGenerate reference images with kie.ai nano-banana-pro, host them on imgbb,and write permanent `image_url` fields back into the JSON files.## Setup: .envOn first run the script automatically creates a `.env` template in your projectdirectory and exits. Open it, fill in your keys, then re-run:```KIE_API_TOKEN=your_kie_api_key_hereIMGBB_API_KEY=your_imgbb_api_key_here```- **KIE_API_TOKEN** — from [kie.ai](https://kie.ai) dashboard → API Keys- **IMGBB_API_KEY** — from [api.imgbb.com](https://api.imgbb.com) (free account)- Add `.env` to `.gitignore` — never commit API keys- Real environment variables always take precedence over `.env` valuesA reference copy is at `assets/.env.example` in this skill's folder.## Prerequisites- `characters.json` and `backgrounds.json` must exist (from Skill 2)- `.env` in project directory with both keys filled in (auto-created on first run)- `pip install requests`## Running the Script```bashcd /path/to/your/projectpython ~/.claude/skills/generating-character-and-background-images/scripts/generate_images.py```### What the script does per image:1. **Generate** — POST to `jobs/createTask` (model: `nano-banana-pro`, aspect_ratio, resolution: `1K`)2. **Poll** — GET `jobs/recordInfo?taskId=...` until `state: success`; extract `resultUrls[0]`3. **Download** — save PNG to `./characters/{id}.png` or `./backgrounds/{id}.png`4. **Upload to imgbb** — POST base64-encoded PNG to `api.imgbb.com/1/upload`; get permanent URL5. **Write back** — store imgbb URL as `image_url` in `characters.json` / `backgrounds.json`Character images: **1:1** aspect ratio (reference sheets)Background images: **16:9** aspect ratio (cinematic wide shots)If imgbb upload fails, the kie.ai URL is used as fallback (may not be permanent).## Regenerating Specific Images1. Update the `prompt` field in the JSON2. Delete the existing PNG (e.g., `./characters/char_001.png`)3. Re-run the script — it skips existing files, regenerates deleted ones## Important: image_url FieldsAfter running, each JSON entry has an `image_url` field containing a permanentimgbb URL. **Skill 5 (`generate_videos.py`) reads these URLs** to generatecomposite images — do not delete or overwrite them.## Review Gate (MANDATORY)After all images are generated, present this EXACTLY:```✅ Image Generation complete.📋 Summary:- Character images: [X] generated → ./characters/ [char_001.png: "Name", char_002.png: "Name", ...]- Background images: [X] generated → ./backgrounds/ [bg_001.png: "Name", bg_002.png: "Name", ...]- Failed: [X] (list any failures)- image_url (imgbb) written to: characters.json, backgrounds.json👉 Please review the generated images. You can: - Approve all → say "approved" or "proceed" - Reject specific images → e.g., "regenerate char_001, the hair is wrong" (update prompt in characters.json, delete the PNG, re-run script) - Replace images manually → save your PNG as ./characters/char_001.png, then tell me to upload it to imgbb to get a fresh image_url⏸️ Waiting for your approval before creating the shot list.```**NEVER** proceed to the next skill automatically. Wait for explicit approval.

Frequently asked questions