GLM-V-Prompt-Gen

Analyze images/videos and generate professional prompts for text-to-image and text-to-video AI tools (Midjourney, Stable Diffusion, DALL-E, Sora, Runway, Kli...

Security Scan

The skill's code, instructions, and requested credential (ZHIPU_API_KEY) are consistent with its stated purpose of analyzing images/videos and generating prompts via the Zhipu/BigModel API; no disproportionate permissions or unrelated endpoints are present, but note that user media and the API key are sent to an external service (privacy consideration).

Purpose & Capability

Name/description (prompt generation from visual inputs) aligns with the code and runtime instructions. The script sends image/video content to BigModel (open.bigmodel.cn) and constructs prompts — requiring a ZHIPU_API_KEY is expected for this integration.

Instruction Scope

Instructions and the script legitimately read local images (or accept URLs/base64), encode/send them to the Zhipu API, and require the full prompt output to be displayed. This is within scope, but the skill will upload user-provided images/videos (including local files) to an external service; users should be aware of privacy implications. The SKILL.md also states the API key is shared with other Zhipu skills (operational note about credential reuse).

Install Mechanism

No install spec; instruction-only with a single Python script. No downloads from untrusted URLs or archive extraction are present. The script uses the requests library (normal for network calls).

Credentials

Only a single required environment variable (ZHIPU_API_KEY) is requested, which matches the external BigModel/Zhipu API usage. The policy note that this key is reused across other Zhipu skills is expected but worth user awareness.

Persistence & Privilege

always is false and the skill is user-invocable. It does not request permanent platform-wide privileges or modify other skills' configs. Normal autonomous invocation (not disabled) is allowed but not combined with other red flags.

Assessment

This skill appears to do what it says: it encodes/local-reads images or accepts URLs and sends them to Zhipu/BigModel to generate prompts. Before installing, consider: (1) Privacy: any local image you provide will be uploaded to an external service — avoid sensitive images. (2) Credential reuse: the ZHIPU_API_KEY will be used by this and other Zhipu skills; store and rotate the key if needed and avoid sharing a production/privileged key. (3) Trust the provider: verify you are comfortable sending your data to bigmodel.cn / open.bigmodel.cn and review their privacy/terms. (4) If you need higher assurance, review the full script locally (it uses only requests and explicit API calls) or run it in an isolated environment. If anything about credential sharing or remote uploads is unacceptable, do not install or use the skill.

GLM-V Prompt Generation Skill

Analyze reference images or videos and generate professional prompts for AI image/video generation tools.

When to Use

• Generate prompts for text-to-image tools (Midjourney, Stable Diffusion, DALL-E, etc.)
• Generate prompts for text-to-video tools (Sora, Runway, Kling, Pika, etc.)
• User mentions "生成prompt", "文生图prompt", "文生视频prompt", "prompt工程", "参考图生成prompt", "generate prompt"
• User provides an image/video and wants to recreate or remix it
• Extract prompt ideas from reference visual content

Supported Input Types

Type	Formats	Max Size	Max Count	Base64
Image	jpg, png, jpeg	5MB / 6000×6000px	50	✅
Video	mp4, mkv, mov	200MB	—	❌ (URL only)

⚠️ Images and videos cannot be used in the same request. ⚠️ Videos only support URLs — local paths and base64 are NOT supported.

📋 Output Display Rules (MANDATORY)

After running the script, you must display the full prompt output exactly as returned. Do not summarize, truncate, or only say "prompt generated". Users need the complete prompt (especially the English prompt) for direct copy/paste.

• Show the full output: content analysis + prompt + prompt breakdown
• In auto mode, show both text-to-image and text-to-video prompts
• English prompts are core output and must be shown completely
• If output was saved (-o), provide the file path and show file content

Output Modes

Mode	Description
image	Generate prompts for text-to-image tools (default)
video	Generate prompts for text-to-video tools
auto	Generate prompts for both image and video

Resource Links

Resource	Link
Get API Key	https://bigmodel.cn/usercenter/proj-mgmt/apikeys
API Docs	Chat Completions / 对话补全

Prerequisites

API Key Setup / API Key 配置（Required / 必需）

This script reads the key from the ZHIPU_API_KEY environment variable and shares it with other Zhipu skills. 脚本通过 ZHIPU_API_KEY 环境变量获取密钥，与其他智谱技能共用同一个 key。

Get Key / 获取 Key： Visit Zhipu Open Platform API Keys / 智谱开放平台 API Keys to create or copy your key.

Setup options / 配置方式（任选一种）：

1. OpenClaw config (recommended) / OpenClaw 配置（推荐）： Set in openclaw.json under skills.entries.glmv-prompt-gen.env:

   "glmv-prompt-gen": { "enabled": true, "env": { "ZHIPU_API_KEY": "你的密钥" } }
   

2. Shell environment variable / Shell 环境变量： Add to ~/.zshrc:

   export ZHIPU_API_KEY="你的密钥"
   

💡 If you already configured another Zhipu skill (for example zhipu-tools or glmv-caption), they share the same ZHIPU_API_KEY, so no extra setup is needed. 💡 如果你已为其他智谱 skill（如 zhipu-tools、glmv-caption）配置过 key，它们共享同一个 ZHIPU_API_KEY，无需重复配置。

How to Use

Image → Text-to-Image Prompt

python scripts/prompt_gen.py --images "https://example.com/photo.jpg"
python scripts/prompt_gen.py --images /path/to/photo.png

Image → Text-to-Video Prompt

python scripts/prompt_gen.py --images "https://example.com/scene.jpg" --mode video

Image → Both (Image + Video Prompts)

python scripts/prompt_gen.py --images "https://example.com/photo.jpg" --mode auto

Video → Text-to-Video Prompt

python scripts/prompt_gen.py --videos "https://example.com/clip.mp4" --mode video

Save Result to File

python scripts/prompt_gen.py --images photo.jpg --mode image -o prompt.md

Custom Model

python scripts/prompt_gen.py --images photo.jpg --model glm-4.6v-flash

Output Example (image mode)

### Content Analysis
A cyberpunk cityscape at night, with dense skyscrapers, glowing neon signs, and rain-wet streets reflecting colorful light.

### Prompt
Cyberpunk cityscape at night, towering skyscrapers with glowing neon signs,
rain-wet streets reflecting colorful lights, flying cars in the distance,
volumetric fog, dramatic lighting, ultra detailed, 8K, cinematic composition

### Prompt Breakdown
- **Subject**: Futuristic skyline with skyscrapers and neon lights
- **Style**: Cyberpunk, sci-fi
- **Color**: Cool/warm contrast with blue-purple dominance and neon accents
- **Lighting**: Neon glow, wet-surface reflections, volumetric fog
- **Composition**: Wide-angle perspective with layered depth
- **Mood**: Mysterious, futuristic, high-tech

CLI Reference

python scripts/prompt_gen.py (--images IMG [IMG...] | --videos VID [VID...]) [OPTIONS]

Parameter	Required	Description
--images, -i	One of	Image paths or URLs (jpg/png/jpeg, base64 OK)
--videos, -v	One of	Video URLs (mp4/mkv/mov, URL only)
--mode, -m	No	Output mode: image (default), video, or auto
--model	No	Model name (default: glm-4.6v)
--temperature, -t	No	Sampling temperature 0-1 (default: 0.6)
--max-tokens	No	Max output tokens (default: 2048)
--thinking	No	Enable thinking/reasoning mode
--stream	No	Enable streaming output
--output, -o	No	Save result to file
--pretty	No	Pretty-print JSON error output

Error Handling

API key not configured: → Guide user to configure ZHIPU_API_KEY

Authentication failed (401/403): → API key invalid/expired → check at Zhipu API Keys / 智谱官网

Rate limit (429): → Quota exhausted → wait and retry

Content filtered: → warning field present → content blocked by safety review

Timeout: → Video processing may take time → increase timeout or use smaller files