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.
npx agentmods add skills/lessweb/deepcode-cli/image-generatornpx skills add lessweb/deepcode-cli --skill image-generatorgit clone --depth 1 https://github.com/lessweb/deepcode-cliWhat 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 | $0.00062 | $0.01525 |
| Opus 5 | $0.00031 | $0.00763 |
| Sonnet 5 | $0.00012 | $0.00305 |
| Haiku 4.5 | $0.00006 | $0.00153 |
Grade A, and why
image-generator 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 yesterday.
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 — 108 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Image Generator
通过 scripts/image_generator.py 试算积分并生成无水印 PNG。支持纯文本生图、多参考图融合，以及依据图片中标记区域进行局部编辑。
Language / 语言
Always respond in the user's language, not the language of this document.
- If the user's latest message or the ongoing conversation is in English (for example
Generate a picture of a baby panda), use English for ALL user-facing text: assistant explanations,AskUserQuestionquestions, option labels and descriptions, confirmation prompts, and final result reports. - If the user writes in Chinese, use Chinese for all of the above.
- When presenting ratio or resolution options, translate the direction labels into the user's language (for example
1:1 Square,3:4 Portrait,4:3 Landscapein English;1:1 正方形,3:4 竖屏,4:3 横屏in Chinese). Do not mix both languages in a single option label.
执行流程
严格按以下顺序执行。不得在用户明确确认积分消耗前调用图片生成接口。
1. 收集真实需求
- 将用户真正想得到的画面整理成完整提示词，保留主体、动作、环境、构图、风格、光线、色彩、文字和限制条件。
- 收集用户提到的全部本地图片路径、HTTP(S) 图片 URL 或 data URL。最多使用 10 张图片，不要把普通网页 URL 当作图片 URL。
- 对多参考图，在提示词中说明每张图承担的角色，例如“参考图 1 的构图、参考图 2 的人物服装、参考图 3 的配色”，避免只写“融合这些图片”。
- 对标记区域编辑，将含标记的图片作为参考图，并在提示词中清楚说明标记的外观、要替换或补充的内容、必须保留的未标记区域，以及最终结果中移除标记并自然融合边缘、透视、光照和纹理。
- 如果需求本身仍有会显著改变结果的歧义，先用
AskUserQuestion补齐，不要擅自添加关键视觉设定。
2. 确认比例和分辨率
如果用户已经明确指定支持的比例或分辨率，直接采用，不要重复询问。如果任一项缺失，必须使用 AskUserQuestion 让用户选择；两项都缺失时，在同一次工具调用中提出两个问题。
比例选项的标签必须带方向符号，不能只显示数字。支持的比例为：
□ 1:1（正方形）▯ 3:4（竖屏）▭ 4:3（横屏）▭ 16:9（横屏）▯ 9:16（竖屏）▯ 2:3（竖屏）▭ 3:2（横屏）▭ 21:9（超宽横屏）
如果 AskUserQuestion 限制选项数量，优先展示最符合需求的若干比例，并在问题中列出其余受支持比例供用户通过“其他”填写。分辨率选项为 1k、1.5k、2k、4k；不要替用户默认选择。
3. 试算积分
从 Skill 目录运行脚本。每张参考图分别传一个 --image：
python3 scripts/image_generator.py cost \
--prompt "<最终提示词>" \
--ratio "<比例>" \
--resolution "<分辨率>" \
--image "<参考图 1>" \
--image "<参考图 2>"
没有参考图时省略全部 --image。脚本会把本地图片编码为 data URL，并直接传递 HTTP(S) URL 或已有 data URL。保存本次调用使用的最终提示词、图片顺序、比例和分辨率，生成时必须原样复用。
4. 请求生成确认
成功取得 credits 后，必须使用 AskUserQuestion 明确询问是否生成。问题中同时写明：
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- yesterday First seen · 108 lines · 62 tokens per session scan A 53568e639dae
image-generator is a skill published in the GitHub repository lessweb/deepcode-cli (2,224 stars, last pushed 4d ago), licensed MIT. It adds 62 tokens to every session and 1,525 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 2026-08-30.
Other skills, from other repositories
codex-router-media
Generate video, music, speech, or images with the operator's MiniMax Token Plan subscription through the codex-router media CLI. Use when the session runs a MiniMax custom (non-OpenAI) model (for example minimax-m3) with the MiniMax Token Plan provider connected, and the user explicitly asks to create a video, a song…
bilibili-helper
B站内容助手 - 视频标题描述优化、标签策略、封面设计建议、分区选择、评论互动.
ppt-designer
PPT 演示助手 - 幻灯片设计、内容结构、视觉优化、动画建议、演讲稿.
douyin-script
抖音/快手短视频脚本 - 前3秒hook、脚本结构、热门音乐建议、话题标签策略.
visual-acceptance
UI/视觉改动交付前的终验方法论——多主题截图矩阵复现、像素真值判据链、CSS 层叠陷阱、布局漂移审查、before/after 存证。当视觉改动需要验收（而非实现）时使用：交付前最后一环，回答「看得见的部分真的对吗」。.
gemini-interactions-api
Use this skill when writing code that calls the Gemini API for text generation, multi-turn chat, multimodal understanding, image generation, video generation, streaming responses, background research tasks, function calling, structured output, or migrating from the old generateContent API. This skill covers the…