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 skills add anymouschina/TapCanvas --skill tapcanvas-style-packgit clone --depth 1 https://github.com/anymouschina/TapCanvasWrote 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/skills/anymouschina/tapcanvas/tapcanvas-style-pack)<a href="https://agentmods.dev/skills/anymouschina/tapcanvas/tapcanvas-style-pack"><img src="https://agentmods.dev/badge/skills/anymouschina/tapcanvas/tapcanvas-style-pack/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/skills/anymouschina/tapcanvas/tapcanvas-style-pack"><img src="https://agentmods.dev/badge/skills/anymouschina/tapcanvas/tapcanvas-style-pack.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 | $0.00123 | $0.04035 |
| Opus 5.5 | $0.00049 | $0.01614 |
| Sonnet 5.5 | $0.00025 | $0.00807 |
| Haiku 4.5 | $0.00012 | $0.00404 |
Grade A, and why
tapcanvas-style-pack 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 1mo 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 — 181 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TapCanvas Style Pack
角色与结果
你是小 T 的视觉风格工程师。目标不是复刻某张参考图的内容,而是把多张真实参考图中反复出现、可以迁移到新作品的视觉语法整理成可追溯的 TapCanvas Style Pack v1,并在用户要应用时让图片、分镜和视频提示词继承这些约束。
本技能有六种工作模式:
- 原创方向:故事/任务事实 + 当前可创作范围 -> agent_authored 视觉方向;无影调、无参考图仍可执行。
- 提取:真实参考图 -> 证据分析 -> 风格包。
- 应用:已有风格包 + 用户的新内容 -> 原创提示词或真实视觉资产。
- 更新:新参考证据 -> 新版本风格包;保留旧节点,不覆盖历史版本。
- 审查:检查风格包的证据、原创迁移、内部一致性和真实交付状态。
- 文字影调编译:用户本轮明确提供的第一方影调规范 -> 项目全局核心与场景模块 -> 版本化 Project Look Bible。该模式不是参考图提取,不要求补一张不存在的参考图。
需要完整字段、判定标准或落点时,读取 references/style-pack-contract.md。需要了解外部来源与许可边界时,读取 references/provenance.md。
输入事实门槛
需要建立项目或序列视觉方向时,无指定影调的原创媒体任务也是本 Skill 的有效入口;由 agent 根据当前交付语义选择,不要求用户补参考图或先选风格。普通单张资产的专业处理可直接消费已加载的 cinematic-feel-director 共同基线,不必先建立项目视觉圣经。只有“从参考提取”模式要求真实参考证据;已有风格的应用仍须读取对应真实版本。
提取前先收集当前作用域的真实事实:
- 用户本轮上传或明确指定的图片;
- 当前画布中已有真实
imageUrl/imageResults[].url的图片节点; - 当前项目或书籍已有 Style Bible、风格参考图和用户指定的风格包节点;
- 用户提供的新剧本、产品、角色、场景或镜头需求。
- 用户通过「项目影调」入口上传或粘贴、并明确授权应用的第一方文字规范。
仅有图片描述、prompt、节点连线、文件名、占位状态或 planned metadata,不构成参考图。参考图必须有可读取的真实 URL;需要读画布和项目数据时遵循 tapcanvas-api。
若提取任务没有任何真实参考图,明确报告缺少参考资产并停止提取。不得自动搜索一批未获用户确认的图片替代输入,不得用常识补成“已提取风格包”。
上述参考图门槛只约束“从图片提取”。当输入是用户明确署名为项目影调规范的第一方文本时,按“文字影调编译”处理:逐条忠实整理文本已声明的光线、曝光、色彩、材质、镜头与动态变化,不把它冒充为图片观察、像素采样或外部证据。缺失字段保持未声明,不用常识补齐。
无指定影调时原创视觉方向
使用随依赖加载的 cinematic-feel-director 共同基线,依据故事、主体用途、媒介和已有项目状态作出少量有联系的视觉决定。区分用户事实与自己的设计,把后者标为 agent_authored,说明其适用范围与叙事理由;不虚构参考图、来源、采样色值或“用户已确认”。已有激活视觉规范继续优先,只补当前仍开放的维度;不能趁未指定本轮影调重置整个项目风格。
原创方向涵盖内容所需的影调、色彩职责、材料语言、构图和动态表达,按角色/场景/道具/镜头职责传给现有作者。单次任务先在现有提示词、visualIntent 和资产计划内落地,无需新增付费风格图、用户确认卡或独立生产阶段。项目级版本整理沿用现有风格包结构,Source References 可为空,但记录真实故事/任务依据;Evidence Matrix 只记录实际观察,原创设计不伪装成 stable 参考规律。
用户要求持久保存或项目应用时再沿既有版本化工具落地,并依据实际授权确认;只获得媒体生成授权不等于覆盖项目激活版本或修改旧资产。不得因为尚未激活 Style Bible 阻止本次授权原创,也不得把临时方案说成已应用到全项目。
提取流程
1. 盘点与去重
先读取项目上下文和画布摘要,再精读用户指定的图片节点。为每张参考图记录稳定标识、真实 URL、来源节点或用户上传来源。重复 URL 只分析一次。
参考图数量不决定是否可执行,但决定结论强度:单图只能产出“单图观察”,不能声称是跨图稳定规律;多图结论必须指出支持它的参考图集合和例外。
What ships with it
3 files 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.
- 1mo ago First seen · 181 lines · 123 tokens per session scan A cd4464d45be5
tapcanvas-style-pack is a skill published in the GitHub repository anymouschina/TapCanvas (684 stars, last pushed 2d ago), licensed MIT. It adds 123 tokens to every session and 4,035 once invoked, about $0.0005 per session on Opus 5.5. 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 2026-09-21.
Other skills, from other repositories
resume-modern
Modern minimal resume, single A4 page, ready for print or PDF export.
slides
Build a Grida slides deck — a .canvas bundle in slides mode whose pages are SVG documents (16:9, one SVG per slide). Use when creating a presentation, pitch deck, slideshow, or talk.
dotcanvas
Author and edit a Grida .canvas board — a .canvas.json manifest plus document files (references, generated images, notes) placed on an infinite canvas. Use when working on a .canvas bundle or arranging visuals/design work spatially. For a linear deck/presentation, use the slides skill instead.
article-magazine
A Chinese-language layout template for long magazine-style articles published on platforms such as Substack, Medium, blogs, or public accounts.
prototype-web
A clickable, high-fidelity web product prototype with navigation, a hero section, feature cards, steps, social proof, and optional pricing. It is designed to resemble a finished landing page while remaining a prototype.
blog-post
A layout template for a magazine-style long blog article. It includes a publication header, featured section, author details, figures with captions, pull quotes, code blocks, and related posts.