据 AlphaMoat 统计,截至 2026-09,Pinterest 累计下载 4,586 次、获 2 个星标,在全部 5,078 个 AI 技能中排名第 2,083 位,在「设计创意」分类(512 个)中排名第 203 位。
| 平台 | 累计安装 | 累计下载 | 星标 | 评论 |
|---|---|---|---|---|
| ClawHub ↗ | 135 | 4,141 | 1 | - |
| SkillHub ↗ | 134 | 4,586 | 2 | — |
各平台的计数口径不同,统计时间不一致。
该技能可查找视觉灵感并分享图片;API路径可读取用户自己的图板和图钉,需要OAuth凭据。外部图片发送前应检查并确认符合预期;令牌和密钥属于敏感凭据,应在共享会话之外保管并限制权限。
资料发布于 2026-10-29
以下是根据技能文档生成的示例,不代表实际运行结果。
2026-09 起统计方法调整:同名技能按作者分别统计,并覆盖全部技能。图中虚线为调整前的月份,与之后的数据不直接比较,也不计算跨越调整的增长。
| 月份 | 累计下载 | 环比 | 星标 |
|---|---|---|---|
| 2026-09 | 4,586 | — | 2 |
| 2026-08 | 4,040 | +4.2% | 2 |
| 2026-07 | 3,878 | +9.0% | 1 |
| 2026-06 | 3,558 | +4.9% | 1 |
| 2026-05 | 3,391 | — | 1 |
以下是作者发布的原始文件,按纯文本显示,本站未做修改或审核;其中的链接不可点击,命令请核实后再运行。抓取于 2026-10-23。
--- name: pinterest description: Search and browse Pinterest pins, get pin details, and send actual images to the user via Telegram/messaging. Use when the user wants to find inspiration, search for images/ideas, or browse Pinterest content. Sends images directly, not just links. --- # Pinterest Skill Search, browse, and share Pinterest pins — sends actual images to chat, not just links. ## Quick Search & Send Images
### Step 1: Search Pinterest
```
browser action=navigate url="https://www.pinterest.com/search/pins/?q=YOUR+SEARCH+TERMS"
browser action=snapshot
```
### Step 2: Get High-Res Image URLs
From the snapshot, find image URLs. Pinterest images follow this pattern:
- Thumbnail: `https://i.pinimg.com/236x/...`
- Medium: `https://i.pinimg.com/564x/...`
- **High-res: `https://i.pinimg.com/originals/...`**
To get high-res: replace `236x` or `564x` with `originals` in the URL.
### Step 3: Send Images to User
**Send actual image (not link!):**
```
message action=send media="https://i.pinimg.com/originals/xx/xx/image.jpg" message="Pin description here"
```
**Send multiple images:**
```
message action=send media="https://i.pinimg.com/originals/..." message="Option 1: Modern minimal"
message action=send media="https://i.pinimg.com/originals/..." message="Option 2: Cozy rustic"
```
## Detailed Pin Workflow
1. **Navigate** to Pinterest search
2. **Snapshot** to see results
3. **Click** on a pin for details (gets larger image)
4. **Screenshot** the pin detail page OR extract originals URL
5. **Send image** via message tool with `media=` parameter
### Getting Original Images
When on a pin detail page:
- Look for `<img>` with `src` containing `i.pinimg.com`
- Convert to originals: `https://i.pinimg.com/originals/{hash}.jpg`
## Example: "Find me minimalist desk setups"
```
# 1. Search
browser action=navigate url="https://www.pinterest.com/search/pins/?q=minimalist+desk+setup"
browser action=snapshot
# 2. Extract image URLs from snapshot (look for i.pinimg.com)
# 3. Convert to high-res originals
# 4. Send images
message action=send media="https://i.pinimg.com/originals/ab/cd/ef123.jpg" message="Clean white desk with plant 🌿"
message action=send media="https://i.pinimg.com/originals/gh/ij/kl456.jpg" message="Wooden desk, natural light ☀️"
```
## Alternative: Screenshot Method
If image URL extraction is tricky, screenshot the pin:
```
browser action=navigate url="https://www.pinterest.com/pin/123456/"
browser action=screenshot
# Then send the screenshot file
message action=send filePath="/path/to/screenshot.jpg" message="Here's the pin!"
```
## API Method (For User's Own Content)
Requires OAuth token setup — see `references/oauth-setup.md`
```bash
export PINTEREST_ACCESS_TOKEN="your_token"
python3 scripts/pinterest_api.py boards
python3 scripts/pinterest_api.py board-pins <board_id>
python3 scripts/pinterest_api.py pin <pin_id>
```
## Key Points
- ✅ **Always send images directly** using `media=` parameter
- ✅ Use `originals` URLs for high-res
- ❌ Don't just send links — send the actual image
- 💡 If URL doesn't work, screenshot the pin and send that
## References
- OAuth setup: `references/oauth-setup.md`
- API endpoints: `references/api-reference.md`
ContextWeave 架构图一键生成
用 Nano Banana Pro 生成或编辑图片的脚本技能。
用 whisper 命令行在本地把音频转写成文字或做翻译的技能。
通过 MCP 服务生成与编辑图表的技能。
面向界面设计的思路与落地建议技能,覆盖布局、交互流程、设计令牌与前端改动。
面向多家视频模型 API 的 AI 视频生成与编排技能。
用 ffmpeg 从视频里取出画面帧,或生成便于快速查看的缩略图。
用本地 Whisper 把音频文件转成文字,模型下载完成后可完全离线运行。
数据月份:2026-09 · 下载/星标/安装数聚合自公开技能注册库(ClawHub、SkillHub),按月更新。口径说明见数据方法。