meitu-image-fix

name	meitu-image-fix
description	自动诊断单张已有图片的画质、人像和内容问题，并按最优顺序串联 image-superres-enhance/image-edit/image-cutout 做综合修复。仅当用户明确要对同一张图做一键综合修复、说不清该用哪个单项工具，或同时提出两个及以上修复诉求时触发；单一明确诉求如仅去水印、仅去背景、仅美颜、仅超清不单独触发。
version	1.0.0
metadata	{"openclaw":{"requires":{"bins":"[Truncated]","env":"[Truncated]","paths":"[Truncated]"},"primaryEnv":"MEITU_OPENAPI_ACCESS_KEY","security":{"dataFlow":"Inputs, selected local context, and generated prompts may be sent to Meitu OpenAPI when used by the workflow.","credentials":"Credentials are used only for CLI authentication and must not be disclosed.","persistence":"Record workflows may access declared project and visual memory/rules files."}}}
security	{"credential_use":"Uses Meitu OpenAPI credentials from env or ~/.meitu/credentials.json for CLI calls; credentials must not be echoed, logged, or embedded in prompts.","remote_processing":"Input images, repair instructions, intermediate URLs, and generated prompts may be sent to Meitu OpenAPI.","persistence":"Project mode may write output files and may update project or visual-memory files according to the Record workflow."}
requirements	{"credentials":[{"name":"MEITU_OPENAPI_ACCESS_KEY","source":"env | ~/.meitu/credentials.json"},{"name":"MEITU_OPENAPI_SECRET_KEY","source":"env | ~/.meitu/credentials.json"}],"permissions":[{"type":"file_read","paths":"[Truncated]"},{"type":"file_write","paths":"[Truncated]"},{"type":"exec","commands":"[Truncated]"}]}

Meitu Image Fix

Overview

接收一张有问题的图片，自动诊断画质/人像/内容三类问题，规划最优修复管线，按顺序调用 meitu-cli 工具链完成修复。核心价值是诊断+串联——用户不需要知道该用哪个工具，只需要说"帮我修一下"。

Dependencies

• meitu-cli (>=2.0.6): npm install -g meitu-cli@latest
• 凭证: 首选 env MEITU_OPENAPI_ACCESS_KEY / MEITU_OPENAPI_SECRET_KEY，或预置 ~/.meitu/credentials.json；仅在用户明确要求写入本地凭证时，再执行 meitu config set-ak / meitu config set-sk

路径别名： 下文中 $VISUAL = {OPENCLAW_HOME}/workspace/visual/

Core Workflow

Preflight → [Context: 跳过] → Execute (诊断 → 规划 → 逐步执行) → Deliver
              ↑ 工具型 skill，无创意上下文需求，始终跳过 Context

Preflight

1. meitu --version → 未安装则提示 npm install -g meitu-cli@latest
2. meitu auth verify --json → 凭证无效则引导配置
3. Detect mode: cwd has openclaw.yaml → project mode; else → one-off can_record = cwd 有 openclaw.yaml AND $VISUAL 存在（两者缺一即 false）
4. Resolve output_dir: openclaw.yaml → ./output/ | else → $VISUAL/output/meitu-image-fix/ mkdir -p {output_dir}

Context

工具型 skill，无创意上下文需求，始终跳过 Context，直接到 Execute。

Execute

诊断：看图 + 听用户

从两个信息源判断问题：

视觉分析（用 Read 工具读取图片，多模态识别问题）：

观察到的现象	标记
分辨率低 / 明显模糊 / 压缩锯齿 / 噪点多	quality:blur
可见水印 / 半透明文字叠印	content:watermark
画面中有不属于主体的路人/杂物	content:unwanted-object
文字内容有错别字	content:text-error
背景有瑕疵（杂乱/破损/穿帮）	content:bg-flaw
人脸可见 + 肤质粗糙/暗沉/痘印明显	portrait:skin
人脸可见 + 五官可优化	portrait:feature

（用户说的话直接映射到标签）：