Psd Automator Screenshot

核心功能：依托psd-automator核心，通过截图结合自然语言指令定位PSD文本层，按要求自动执行修改，设置信度校验——高置信请求自动执行，中置信返回候选待确认。典型场景：快速修改PSD海报、批量调整设计文案等。输入为含agentId、截图及修改需求的指定格式指令，优先用/main-image-editor，兼容旧指令；输出为修改结果或待确认候选，所有调用记录至skills-usage.json。

dhrxy

热度 66.38

最近更新 2026-05-11 17:29:05

分类：自动化依赖：MCP 服务 / 读写文件应用场景：自动化脚本 / 工作流

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add psd-automator-screenshot

Metadata

name: psd-automator-screenshot
description: Use screenshot + natural language instruction to locate PSD text layers and dispatch automated edits with confidence gating. Requires psd-automator core.
metadata: {"openclaw":{"userInvocable":true,"commandDispatch":"tool","commandTool":"psd_automator_screenshot","commandArgMode":"raw"}}

psd-automator-screenshot

This skill dispatches screenshot-driven PSD text updates while keeping the existing psd-automator execution pipeline.

Preferred command

Use the built-in command first:

/main-image-editor <agentId> <中文需求> --screenshot ~/Desktop/修改3.png

Use /psd-automator-screenshot only when you need backward-compatible command behavior.

Usage

/psd-automator-screenshot <agentId> <截图改字指令>

Example:

/psd-automator-screenshot design-mac-01 找到banner.psd或banner.psb，将红框区域替换成"限时199元"，保持字体和字号不变

Notes

High confidence requests are auto-dispatched.
Medium confidence requests return top candidates for confirmation.
Every invocation/execution is recorded to skills-usage.json.