GitHub
公开源码、更新记录与 Issue 都在这里。
一键抓取品牌 logo、标准化处理并接入官网 partners 区块,多语言全覆盖
基本信息
Maintain the Skill Publisher website's partners section AND align partner logo rows on event posters / hero strips: reuse lov-find-logo for brand logo discovery, normalize collected logos to a 240px-tall content canvas (retina-ready), rasterize SVGs via rsvg-convert before normalizing (so SVG viewBox padding gets cropped), strip embedded background rects from icon-style SVGs, composite icon + wordmark when only an icon is available (using brand fonts), wrap logos in a fixed-size grid box (96×30 with subtle border) for stable matrix layouts, replace existing logos with user-provided files, append new partners to the PARTNERS array with i18n taglines across zh-CN/en/ja/th, and audit the section for dead URLs / missing files / missing translations. Also handles cross-asset visual height parity (multi-logo strips on dark backgrounds, "logo 不等高", unified-color filter recipe). Trigger when the user mentions "合作伙伴", "partners", "trusted by", "新增 logo", "标准化 logo", "替换 logo", "审计合作伙伴", "维护合作伙伴", "logo 不一样高", "logo 对齐", "logo 大小不一致", "logo 颜色不统一", "logo 不清晰", "logo 糊了", "矩阵格子", "等宽 box", "图标加文字", "compose wordmark".
SKILL CARD / TRUST RECORD
这个 Skill 尚未发布完整的可信度卡。案例、维度地图与定价依据会在补齐后显示。
PRICING CARD
复用 find-logo 收集品牌 logo,标准化处理并接入官网 partners 区块,多语言全覆盖。
免费入口
免费
稳定价格带 免费
价值锚点
一次结果通常可减少 1–6 小时的试稿、排版或资产整理。
定价依据
按可预览、可导出、可继续编辑的视觉结果定价,价值来自减少反复试稿。
使用边界
结果依赖输入素材、品牌约束和外部模型;复杂定制、素材采购与线下制作另计。
维护说明
随渲染工具、字体、模型和品牌资产变化复评;保持输出结构比追逐单一风格更重要。
使用说明
Maintain the Skill Publisher website's "Trusted By" partners section: collect brand
logos through lov-find-logo, normalize to the 80px canvas, append
entries with i18n taglines across 4 locales, and audit for dead URLs /
missing assets.
Part of skills — by example.com
SKILLS_DIR="${SKILL_SKILLS_INSTALL_DIR:?Set SKILL_SKILLS_INSTALL_DIR}"
git clone https://example.com/skills/maintain-partners-skill "$SKILLS_DIR/lov-maintain-partners"
git clone https://example.com/skills/find-logo-skill "$SKILLS_DIR/lov-find-logo"
python3 -m pip install Pillow
brew install librsvg # for SVG logo sources
Set the website repo root with --repo, SKILL_MAINTAIN_PARTNERS_SITE_ROOT, or the shared
profile at ${SKILL_PROFILE_PATH:-$HOME/.skill-publisher/skills/profile.json}.
SKILL_WEB_ROOT and PARTNERS_SITE_ROOT remain accepted as legacy aliases.
Set the PARTNERS TSX file with --partners-file, SKILL_MAINTAIN_PARTNERS_FILE, or profile
keys. The default checks app/(main)/(home)/PartnersGrid.tsx first, then
legacy app/(main)/(home)/WorkshopDispatch.tsx.
SKILL_PARTNERS_FILE and PARTNERS_FILE remain accepted as legacy aliases.
Supported profile keys:
{
"sites": {
"skill-publisher_web": "$HOME/projects/my-site",
"partners_file": "app/(main)/(home)/PartnersGrid.tsx"
},
"skill-publisher": {
"web_root": "$HOME/projects/my-site",
"partners_file": "app/(main)/(home)/PartnersGrid.tsx"
},
"workspace": {
"web_root": "$HOME/projects/my-site",
"partners_file": "app/(main)/(home)/PartnersGrid.tsx"
}
}
The Skill Publisher homepage runs a "Trusted By" strip that renders 30+ partner
logos against a grayscale opacity-60 filter. Maintaining it means three
recurring tasks:
lov-find-logo to pull and archive brand logos.PARTNERS in the configured partners
TSX file and adding a partner*Tagline key to all 4 locale JSONs
(zh-CN / en / ja / th).Logo discovery is delegated to lov-find-logo; this skill does not keep
its own homepage crawler or fallback scraper.
This skill is three single-file Python CLIs plus an AI workflow that orchestrates them.
normalize_logo.py Trim, optional inversion, resize to 80px, write PNG
add_partner.py Append to PARTNERS array + i18n JSONs (idempotent)
audit_partners.py Walk PARTNERS; report missing logos / i18n keys / dead URLs
# Collect a logo through the required find-logo skill
SKILL_ROOT="${SKILL_SKILLS_INSTALL_DIR:?Set SKILL_SKILLS_INSTALL_DIR}"
WEB_ROOT="${SKILL_MAINTAIN_PARTNERS_SITE_ROOT:?Set this or pass --repo}"
PARTNERS_TSX="${SKILL_MAINTAIN_PARTNERS_FILE:-app/(main)/(home)/PartnersGrid.tsx}"
python3 "$SKILL_ROOT/lov-find-logo/scripts/find_logo.py" \
--name "Example" --url https://example.com --slug example --json
# Normalize: auto-invert white-on-transparent
normalize_logo.py --src ~/.skill-publisher/logo-collection/example/logo.png \
--dst "$WEB_ROOT/public/partners/example/logo.png"
# Add to PARTNERS + i18n
add_partner.py --repo "$WEB_ROOT" \
--partners-file "$PARTNERS_TSX" \
--name "Example" --href "https://example.com" \
--logo "/partners/example/logo.png" \
--key partnerExampleTagline \
--category community \
--zh "Example · 一句话定位" \
--en "Example · one-line positioning" \
--ja "Example · 一行紹介" \
--th "Example · บรรยายหนึ่งบรรทัด"
# Audit (use --probe to HTTP-check every URL)
audit_partners.py --repo "$WEB_ROOT" --partners-file "$PARTNERS_TSX" --probe
<web-root>/
├── app/(main)/(home)/PartnersGrid.tsx ← PARTNERS array
├── public/partners/<slug>/logo.png ← logo files
└── src/i18n/messages/
├── zh-CN.json ← dispatch.partner*Tagline
├── en.json
├── ja.json
└── th.json
If you fork this for another site, edit the constants at the top of each script.
MIT
登录后即可留言
安装渠道
Yoda、Codex、Claude Code 和其他兼容运行环境都使用同一份 Skill;选择你的工具后,按对应步骤完成安装。
先保存一次工作资料,支持 Profile 的 Skill 会在运行时直接复用。
配置工作资料安装与配置
适合同时使用多个 Agent 的工作流:在一个地方完成安装、分配和调用。
在 Yoda 内完成
Yoda 会把安装后的 Skill 同步给已配置的 Agent,无需手动维护多个目录。
第一次调用
使用 maintain-partners 完成以下任务,并按 Skill 的验收标准检查结果:该 Skill 声明:Requires the lov-find-logo skill plus Python 3.8+ with Pillow (`pip install Pillow --break-system-packages`). Requires rsvg-convert (`brew install librsvg`) when the selected logo source is SVG. Tested on macOS; Linux should work. Website repo paths are configurable via --repo, SKILL_MAINTAIN_PARTNERS_SITE_ROOT, or the shared user profile; this skill must not require Mark's personal absolute path. Legacy path aliases remain accepted for existing local setups.
搭配使用
更多信息
NEXT STEP
回到目录,从结果、证据、依赖和价格选择最合适的工作流。
LOVSTUDIO / KEEP MAKING