Tapcanvas Replicate
Documentstapcanvas-replicateTapcanvas Replicate is an Agent Skill for documents published in the anymouschina/TapCanvas repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
多资产复刻能力：基于 assetInputs（N 张图）执行角色或主体替换，保持版式与镜头连续性。
QUICK START
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/anymouschina/TapCanvas/blob/HEAD/apps/agents-cli/skills/tapcanvas-replicate/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/tapcanvas-replicate, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
TapCanvas Replicate
你负责处理“复刻/替换”任务，输入通常来自 assetInputs，不固定为两张图。
输入理解
role=target：被改造目标图（优先保持构图、镜头、版式）。role=reference/character/product/style/context/mask：参考约束图。- 若存在多个参考图，先提取一致性锚点（形状比例、材质、配色、关键识别特征）。
执行策略
- 先确认目标图与参考图集合。
- 输出替换策略摘要：保留项、替换项、冲突项。
- 对 9 宫格/分镜任务：保持分格与镜头顺序，逐格替换主体身份。
- 对带版式与文字区的设计图：保持信息层级和文案区可读性，替换主体并维持原有表达结构。
失败策略（强制）
- 关键资产缺失、冲突无法解时，直接报错并说明具体缺口。
- 禁止静默兜底、禁止编造未在输入中出现的关键识别特征。