GitHub
公开源码、更新记录与 Issue 都在这里。
按品牌名或网址抓取 logo,自动评分择优(偏好长条形 + 透明底),统一归档到本地,方便网站/PPT/海报罗列
基本信息
Fetch a company/product logo from public sources (Clearbit, og:image, favicon) given a brand name or URL, score candidates (wide-aspect + transparent preferred), and archive the best + runner-ups to the configured logo collection directory. Trigger when the user says "find logo", "找 logo", "抓 logo", "收集 logo", "brand asset", "需要 <brand> 的 logo", or wants logos laid out for a website/PPT/poster.
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
按品牌名或网址抓取 logo,自动评分择优(偏好长条形 + 透明底),统一归档到本地,方便网站/PPT/海报罗列。
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
Fetch a brand/product logo from public sources (Clearbit, og:image, favicon),
score candidates (prefer wide-aspect + transparent), and archive the winner
plus alternates under ~/.skill-publisher/logo-collection/<slug>/.
Useful for building website partner strips, PPT footer rows, poster credits — anywhere you need a lineup of logos that look consistent.
Part of skills — by example.com
npx skills add find-logo -g -y
Stdlib only — no pip install needed.
# by name
python3 scripts/find_logo.py --name "Anthropic"
# by URL (more reliable for non-.com or ambiguous brands)
python3 scripts/find_logo.py --name "xAI" --url https://x.ai
# machine-readable output for chaining
python3 scripts/find_logo.py --url https://stripe.com --json
| Option | Default | Description |
|---|---|---|
--name | — | Brand/product name. Used for slug + meta. |
--url | — | Official URL or bare domain. Overrides the name-based guess. |
--slug | slugified name | Override the archive directory slug. |
--out | SKILL_FIND_LOGO_OUTPUT_DIR or ~/.skill-publisher/logo-collection | Archive root. |
--keep-alts | 2 | How many runner-up candidates to keep. |
--json | off | Emit JSON to stdout. |
At least one of --name or --url is required.
~/.skill-publisher/logo-collection/<slug>/
├── logo.<ext> # highest-scoring candidate
├── alt-1.<ext> # runner-ups (count = --keep-alts)
├── alt-2.<ext>
└── meta.json # brand, source URL per candidate, dims, format, alpha, score
Higher is better. Intended to bubble up wide-aspect + transparent logos so they line up cleanly next to each other.
| Factor | Points |
|---|---|
| Format SVG / PNG / WebP / JPG / ICO | +40 / +20 / +10 / -10 / -20 |
| Has alpha (transparent) | +30 |
| Aspect ≥ 2:1 (banner) | +25 |
| Aspect ≥ 1.3:1 (landscape) | +10 |
| Aspect ≈ 1:1 (square) | -5 |
| Aspect tall / portrait | -15 |
| Short edge ≥ 128 / ≥ 64 / < 32 px | +15 / +5 / -20 |
| Payload < 400 bytes (stub) | -30 |
https://logo.clearbit.com/<domain> (unauthenticated tier).<link rel="icon"> — scraped from the brand's homepage. Usually wins because sites ship a proper share card.https://www.google.com/s2/favicons?domain=<domain>&sz=256 as a safety net.If all three fail, the script exits 2 with status: "no-candidates". The
calling agent is expected to fall back to a web search for a press-kit page.
Archive files default to ~/.skill-publisher/logo-collection/. Override per call with
--out, or set SKILL_FIND_LOGO_OUTPUT_DIR.
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 find-logo 完成以下任务,并按 Skill 的验收标准检查结果:该 Skill 声明:Requires Python 3.8+ (stdlib only — no pip deps). Cross-platform: macOS, Windows, Linux.
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