AgentStack
SKILL verified MIT Self-run

Image Generate

skill-cuteweather-agent-image-generate · by cuteweather

根据提示词进行文生图并保存到本地,支持比例、尺寸、数量、seed、prompt 优化和风格参数。当用户提到“文生图/生成图片/image-01/image-01-live”时使用。

No reviews yet
0 installs
2 views
0.0% view→install

Install

$ agentstack add skill-cuteweather-agent-image-generate

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

Are you the author of Image Generate? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

文生图

目标

  • 根据文本提示词生成图片并保存到本地目录。
  • 返回生成任务信息与本地文件路径。
  • 默认使用独立脚本,不在对话里拼长请求体。

API 配置

  • Endpoint: 由脚本内 API_URL 定义
  • Header: 由脚本内鉴权逻辑自动处理
  • 默认模型:image-01

关键说明

  • response_format=url 时,接口返回临时图片 URL,脚本会立即下载到本地。
  • response_format=base64 时,脚本会直接解码并保存为本地 PNG。

参数约定

  1. 必须传 prompt
  2. 常用参数:modelaspect_rationseedprompt_optimizer
  3. 如果同时传 width/height,需要两个都提供。
  4. n 取值范围为 1-9
  5. image-01-live 可配合 style_type/style_weight 使用。

输出约定

  • 默认输出目录由调用方给定。
  • 文件名默认使用 prompt slug,加三位序号。
  • 可选 --metadata-file 保存完整返回信息。
  • 默认只打印一条汇总 JSON;如需逐张打印,显式使用 --print-each

工具脚本(独立文件)

  • 仓库内脚本路径:image-generate/scripts/text_to_image.py
  • 环境变量:MINIMAX_API_KEY
  • 可选环境变量:MINIMAX_API_URL
SCRIPT="./image-generate/scripts/text_to_image.py"
[ -f "$SCRIPT" ] || { echo "text_to_image script not found"; exit 1; }

可执行示例

示例 1:基础文生图

SCRIPT="./image-generate/scripts/text_to_image.py"
python "$SCRIPT" \
  --prompt "a serene mountain landscape at sunrise, photorealistic" \
  --output-dir "./assets/t2i" \
  --aspect-ratio "16:9" \
  --n 1

示例 2:固定 seed,生成 4 张

SCRIPT="./image-generate/scripts/text_to_image.py"
python "$SCRIPT" \
  --prompt "a cozy reading room with warm lamp light, cinematic interior photography" \
  --output-dir "./assets/t2i" \
  --aspect-ratio "4:3" \
  --seed 42 \
  --n 4 \
  --metadata-file "./assets/t2i/reading_room.json"

示例 3:开启 prompt 优化

SCRIPT="./image-generate/scripts/text_to_image.py"
python "$SCRIPT" \
  --prompt "futuristic city street at night with neon reflections on wet pavement" \
  --output-dir "./assets/t2i" \
  --aspect-ratio "16:9" \
  --prompt-optimizer \
  --n 2

示例 4:image-01-live 风格化生成

SCRIPT="./image-generate/scripts/text_to_image.py"
python "$SCRIPT" \
  --prompt "a small wizard walking through an enchanted forest" \
  --output-dir "./assets/t2i" \
  --model "image-01-live" \
  --style-type "水彩" \
  --style-weight 0.8 \
  --n 1

执行策略(给模型)

  • 需要真实生成图片时优先使用本技能,不要返回占位图或仅给伪代码。
  • 先确认脚本存在,再执行。
  • 若返回鉴权失败、余额不足或参数错误,直接把 status_code/status_msg 回报给用户。
  • 若用户要求稳定复现,优先设置 seed

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Versions

  • v0.1.0 Imported from the upstream source.