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/damndager/image2-studio/image2npx skills add DAMNDAGER/Image2-Studio --skill image2git clone --depth 1 https://github.com/DAMNDAGER/Image2-StudioWhat 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 | $0.00057 | $0.00620 |
| Opus 5 | $0.00028 | $0.00310 |
| Sonnet 5 | $0.00011 | $0.00124 |
| Haiku 4.5 | $0.00006 | $0.00062 |
Grade A, and why
image2 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 2d 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 — 30 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Local Image2
Use the local Image2 service as the primary image-generation path. Do not use the built-in imagegen tool.
Workflow
- Optimize the user's prompt while preserving intent. Make subject, action, composition, viewpoint, style/material, lighting, palette, aspect ratio, and constraints explicit. Do not invent text, logos, people, or objects. For edits, state what must be preserved and what must change.
- Prefer the installed global CLI when available:
%CODEX_HOME%\skills\image2\Image2CLI.exe --prompt "...". The installer places a non-secretimage2-client.jsonbeside it so the CLI knows the service, token, output directory, and launcher. If the packaged CLI is unavailable, use the projectimage2_cli.pywith its absolute path. - The local backend uses the active provider selected in the HTML console. There is no default provider; if setup is incomplete, direct the user to the local API settings instead of attempting a real request.
- If the local service is unavailable, let the configured CLI launcher start it and retry once. For a restricted Codex sandbox, request one host-approved network-enabled execution for the service process before a real provider request. Keep the service on
127.0.0.1. - For a public reference image URL, add
--image-url "https://...". For a local attachment, add--image-path "C:\path\to\image.png", repeating it for multiple images up to eight. - Inspect the CLI JSON response, read the returned local output path, and include the generated image in the conversation. Briefly report the optimized prompt and active provider.
Simulation
For a requested safety check or dry run, use --dry-run. Confirm the active provider model, route, request body mode, size, and quality, then stop before the final provider request. A dry run must not call an external provider or consume image quota.
Security
- Never print, copy, or place an API key in HTML, JavaScript, prompts, shell arguments, logs, or conversation output.
- Provider keys are entered through the local console and stored by the backend in the operating-system credential store. The frontend only receives a configured/not-configured flag.
- Keep the service bound to
127.0.0.1; never expose it on0.0.0.0. - Treat custom provider URLs and reference image URLs as untrusted input. Use the backend validation and do not bypass its limits.
- If a provider reports an outbound-network failure, stop retrying in the restricted sandbox, request one approved network-enabled service execution, and retry only once after approval.
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.
- 2d ago First seen · 30 lines · 57 tokens per session scan A 0029e0661b5a
image2 is a skill published in the GitHub repository DAMNDAGER/Image2-Studio (3 stars, last pushed 11d ago), licensed MIT. It adds 57 tokens to every session and 620 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 skills, from other repositories
xiaohongshu-title
《小红书运营手册 · AI工作台》标题生成与优化技能。用于用户明确要为小红书笔记、封面短句、图文/短视频标题、种草、教程、测评、服务内容起标题、改标题、做标题方向选择时触发。输入可以是正文草稿、选题方向、产品卖点、账号定位、图片/视频描述或已有标题。默认用快速模式输出标题;用户要求分析、首推、理由、方向、诊断或优化时用诊断/优化模式。不要在普通文章标题、论文标题、新闻标题、课程命名等非小红书/社媒场景中自动触发。.
xiaohongshu-magazine
《小红书运营手册 · AI工作台》杂志感选题库技能。用于用户想把小红书账号当成一本杂志来办、确定内容母题(刊魂)、设计固定栏目、用开枝散叶方法批量生成选题、区分账号围绕业务还是围绕个人 IP、让每一篇都在攒信任或攒喜欢、或搭建可长期复用的选题库时触发。输入可以是账号定位、目标用户、产品或服务、个人经历和已有素材。输出母题、栏目结构、按信任和喜欢分类的选题库、每根栏目的角度公式和优先级建议。它负责战略层选题库,日历和系列排期交给 xiaohongshu-topic-planner。.
xiaohongshu-suite
《小红书运营手册 · AI工作台》母技能。用于用户想整体优化小红书账号、内容、标题、主页、选题、评论区或成交路径,但不确定该用哪个具体技能时触发。负责判断问题类型、推荐子技能、组织组合工作流。不要替代子技能做完整输出;当用户只要标题、主页、母题栏目、选题、评论回复或成交路径时,直接切换或引导使用对应 xiaohongshu- 子技能。.
xiaohongshu-comment-reply
《小红书运营手册 · AI工作台》评论区回复技能。用于用户想回复小红书评论、生成置顶评论、引导讨论、处理质疑、引导私信、维护账号人设或批量回复评论时触发。输入可以是笔记主题、账号语气、评论列表、是否希望转化和回复边界。输出多种语气回复、置顶评论和风险提醒。.
xiaohongshu-conversion-path
《小红书运营手册 · AI工作台》成交路径设计技能。用于用户想把小红书内容、主页承接、置顶笔记、评论区、私信和产品/服务转化串起来时触发,重点处理从刷到内容到私信、预约、购买、体验、反馈或成交的路径。适合服务号、产品号、个人 IP、自由职业者和小程序推广。输入可以是账号定位、产品/服务、客单价、用户顾虑、当前内容和转化目标。输出内容分工、信任路径、私信筛选和成交动作。若用户只想改简介、看主页第一眼或优化置顶表达,应使用 xiaohongshu-profile。.
xiaohongshu-profile
《小红书运营手册 · AI工作台》主页体检和简介改写技能。用于用户想优化主页第一眼、账号简介、昵称、头像描述、置顶笔记、信任感和定位表达时触发。输入可以是主页截图描述、昵称、简介、置顶笔记、账号定位、产品/服务和目标用户。输出主页诊断、问题优先级、简介改写和置顶建议。若用户重点询问内容到私信、购买、预约、体验或成交路径,应使用 xiaohongshu-conversion-path。.