Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/an8079/take-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/an8079/take-skills/takes-imapo)<a href="https://agentmods.dev/commands/an8079/take-skills/takes-imapo"><img src="https://agentmods.dev/badge/commands/an8079/take-skills/takes-imapo/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/an8079/take-skills/takes-imapo"><img src="https://agentmods.dev/badge/commands/an8079/take-skills/takes-imapo.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00050 | $0.01466 |
| Opus 5 | $0.00025 | $0.00733 |
| Sonnet 5 | $0.00010 | $0.00293 |
| Haiku 4.5 | $0.00005 | $0.00147 |
Grade A, and why
takes-imapo scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 144 lines — stays where its author put it; the contents beside it link to each section on GitHub.
📷 /takes-imapo — Image Prompt Engineer
AI 图像提示词工程师。将视觉概念转化为精确的提示词，生成专业级摄影质量的图像。
提示词结构框架
主体描述层
- 主主体：主要焦点详细描述（人物、物体、场景）
- 主体细节：属性、表情、姿势、纹理、材质
- 主体与环境关系：与环境或其他元素的互动
- 比例与尺度：大小关系和空间位置
环境与场景层
- 场景类型：影棚、户外、城市、自然、室内、抽象
- 环境细节：特定元素、纹理、天气、时间
- 背景处理：清晰、模糊、渐变、极简
- 大气条件：雾、雨、尘、霾
光线规格层
- 光源：自然光（黄金时刻、阴天、阳光）或人造光（柔光箱、轮廓灯、霓虹）
- 光线方向：正面、侧面、背面、顶光、伦勃朗光、蝶光
- 光线质量：硬光/柔光、散射、镜面、体积光、戏剧光
- 色温：暖色调、冷色调、中性、混合光源
摄影技术层
- 相机视角：平视、低角度、高角度、鸟瞰、虫视
- 焦距效果：广角畸变、长焦压缩、标准镜头
- 景深：浅景深（人像）、深景深（风景）、选择性对焦
- 曝光风格：高调、低调、平衡、HDR、剪影
风格与美学层
- 摄影类型：人像、时尚、编辑、商业、纪实、艺术
- 时代风格：复古、当代、复古、未来主义、永恒
- 后期处理：胶片模拟、调色、对比度处理、颗粒感
- 参考摄影师：风格影响（Annie Leibovitz、Peter Lindbergh 等）
摄影类型模板
人像摄影
[主体描述：年龄、种族、表情、服装] |
[姿势和肢体语言] |
[背景处理] |
[光线设置：主光、补光、轮廓光、头发光] |
[相机：85mm f/1.4 平视] |
[风格：编辑/时尚/商业/艺术] |
[调色板和氛围] |
[参考摄影师风格]
产品摄影
[产品描述和材质细节] |
[表面/背景描述] |
[光线：柔光箱位置、反光板、渐变] |
[相机：微距/标准、角度、距离] |
[主图/生活方式/细节/比例场景] |
[品牌美学对齐] |
[后期处理：干净/氛围/生动]
风景摄影
[地点和地质特征] |
[时间和大气条件] |
[天气和天空处理] |
[前景、中景、背景元素] |
[相机：广角、深焦、全景] |
[光线质量和方向] |
[调色板：自然/增强/戏剧性] |
[风格：纪实/艺术/空灵]
平台优化
| 平台 | 语法特点 |
|---|---|
| Midjourney | 参数 --ar, --v, --style, --chaos，多提示权重 |
| DALL-E | 自然语言优化，风格混合 |
| Stable Diffusion | Token 权重、embedding 引用、LoRA 集成 |
| Flux | 详细自然语言描述，写实强调 |
触发方式
/takes-imapo # 激活图像提示词工程模式
/takes-imapo [视觉概念描述] # 直接生成提示词
帮我写一个[类型]的提示词 # 自然语言触发
输出示例
用户输入：
帮我生成一个电影感的人像照片提示词
优化输出：
## 电影感人像摄影提示词
[主体]一位 30 岁左右的亚洲男性，穿着深色西装，表情沉静而深邃，
眼神中带着故事感。
[光线]电影级戏剧性布光：45 度侧光从相机左侧打来，
形成伦勃朗三角，轮廓光勾勒出面部线条，
淡淡的补光控制阴影深度。
[相机]85mm f/1.4 镜头，平视角度，
浅景深，奶油般焦外虚化。
[环境]深色背景，略带纹理，
可能有烟雾增添氛围感。
[风格]电影海报质感，
参考 Christopher Nolan 电影的视觉语言，
8K 分辨率，电影级调色。
## 完整提示词（Midjourney 格式）
Dramatic cinematic portrait of a 30-year-old Asian male in dark suit, deep contemplative expression with storytelling eyes, cinematic lighting setup: strong key light from 45 degrees creating Rembrandt triangle on face, subtle fill, rim light defining jawline, light atmospheric haze, shot on 85mm f/1.4 at eye level, shallow depth of field with creamy bokeh, dark textured background with subtle gradient, film grain, inspired by Christopher Nolan cinematography, 8k resolution, cinematic color grading --ar 3:4 --style raw --v 6.1
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 144 lines · 50 tokens per session scan A 7273edf200a2
takes-imapo is a command published in the GitHub repository an8079/take-skills (4 stars, last pushed 5mo ago), licensed MIT. It adds 50 tokens to every session and 1,466 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
page-serve
Hand a human a rendered HTML page at a stable HTTPS URL. Registers a portless route for a file or directory (https://NAME.localhost/, never a :port), prints the URL, optionally screenshots it through agent-browser, and tears it down on stop. Use whenever an agent has produced a page a human is meant to open: glyph…
responsive-audit
Drive Playwright MCP to load live URLs at multiple viewport widths and report responsive breakage static review can't see.
scaffold
Scaffold a new project from a Starter Series template (Discord/Telegram bot, MCP server, browser/VS Code extension, Electron, React Native, Cloudflare Pages, npm package, Docker deploy).
frontend-dom
You are an expert in DOM manipulation, browser APIs, Web Components, and advanced client-side engineering.
add-i18n
Implement multi-language support using next-intl. Set up locale detection from browser, URL, or user preference. Create translation files in JSON format for each locale (en, es, fr, etc.). Implement locale-based routing with Next.js App Router. Create language switcher component with dropdown or flags. Translate all…
add-i18n
Implement multi-language support using next-intl. Set up locale detection from browser, URL, or user preference. Create translation files in JSON format for each locale (en, es, fr, etc.). Implement locale-based routing with Next.js App Router. Create language switcher component with dropdown or flags. Translate all…