\r \r

Felo Web Extract Skill\r

\r

When to Use\r

\r Trigger this skill when the user wants to:\r \r

• Extract or scrape content from a webpage URL\r
• Get article/main text from a link\r
• Convert a webpage to Markdown or plain text\r
• Capture readable content from a URL for summarization or processing\r \r Trigger keywords (examples):\r \r
• extract webpage, scrape URL, fetch page content, web extract, url to markdown\r
• Explicit: /felo-web-extract, "use felo web extract"\r
• Same intent in other languages (e.g. 网页抓取, 提取网页内容) also triggers this skill\r \r Do NOT use for:\r \r
• Real-time search or Q&A (use felo-search)\r
• Generating slides (use felo-slides)\r
• Local file content (read files directly)\r \r

Setup\r

\r

1. Get API key\r

\r

1. Visit felo.ai\r
2. Open Settings -> API Keys\r
3. Create and copy your API key\r \r

2. Configure environment variable\r

\r Linux/macOS:\r \r

export FELO_API_KEY="your-api-key-here"\r
```\r
\r
Windows PowerShell:\r
\r
```powershell\r
$env:FELO_API_KEY="your-api-key-here"\r
```\r
\r
## How to Execute\r
\r
### Option A: Use the bundled script or packaged CLI\r
\r
**Script** (from repo):\r
\r
```bash\r
node felo-web-extract/scripts/run_web_extract.mjs --url "https://example.com/article" [options]\r
```\r
\r
**Packaged CLI** (after `npm install -g felo-ai`): same options, with short forms allowed:\r
\r
```bash\r
felo web-extract -u "https://example.com" [options]\r
# Short forms: -u (url), -f (format), -t (timeout, seconds), -j (json)\r
```\r
\r
Options:\r
\r
| Option | Default | Description |\r
|--------|---------|-------------|\r
| `--url` | (required) | Webpage URL to extract |\r
| `--format` | markdown | Output format: `html`, `text`, `markdown` |\r
| `--target-selector` | - | CSS selector: extract only this element (e.g. `article.main`, `#content`) |\r
| `--wait-for-selector` | - | Wait for this selector before extracting (e.g. dynamic content) |\r
| `--readability` | false | Enable readability processing (main content only) |\r
| `--crawl-mode` | fast | `fast` or `fine` |\r
| `--timeout` | 60000 (script) / 60 (CLI) | Request timeout: script uses **milliseconds**, CLI uses **seconds** (e.g. `-t 90`) |\r
| `--json` / `-j` | false | Print full API response as JSON |\r
\r
### How to write instructions (target_selector + output_format)\r
\r
When the user wants a **specific part** of the page or a **specific output format**, phrase the command like this:\r
\r
- **Output format**: "Extract as **text**" / "Get **markdown**" / "Return **html**" → use `--format text`, `--format markdown`, or `--format html`.\r
- **Target one element**: "Only the **main article**" / "Just the **content inside** `#main`" / "Extract only **article.main-content**" → use `--target-selector "article.main"` or the selector they give (e.g. `#main`, `.main-content`, `article .post`).\r
\r
Examples of user intents and equivalent commands:\r
\r
| User intent | Command |\r
|-------------|---------|\r
| "Extract this page as plain text" | `--url "..." --format text` |\r
| "Get only the main content area" | `--url "..." --target-selector "main"` or `article` |\r
| "Extract the div with id=content as markdown" | `--url "..." --target-selector "#content" --format markdown` |\r
| "Just the article body, as HTML" | `--url "..." --target-selector "article .body" --format html` |\r
\r
Examples:\r
\r
```bash\r
# Basic: extract as Markdown\r
node felo-web-extract/scripts/run_web_extract.mjs --url "https://example.com"\r
\r
# Article-style with readability\r
node felo-web-extract/scripts/run_web_extract.mjs --url "https://example.com/article" --readability --format markdown\r
\r
# Raw HTML\r
node felo-web-extract/scripts/run_web_extract.mjs --url "https://example.com" --format html --json\r
\r
# Only the element matching a CSS selector (e.g. main article)\r
node felo-web-extract/scripts/run_web_extract.mjs --url "https://example.com" --target-selector "article.main" --format markdown\r
\r
# Specific output format + target selector\r
node felo-web-extract/scripts/run_web_extract.mjs --url "https://example.com" --target-selector "#content" --format text\r
```\r
\r
### Option B: Call API with curl\r
\r
```bash\r
curl -X POST "https://openapi.felo.ai/v2/web/extract" \\r
  -H "Authorization: Bearer $FELO_API_KEY" \\r
  -H "Content-Type: application/json" \\r
  -d '{"url": "https://example.com", "output_format": "markdown", "with_readability": true}'\r
```\r
\r
## API Reference (summary)\r
\r
- **Endpoint**: `POST /v2/web/extract`\r
- **Base URL**: `https://openapi.felo.ai`. Override with `FELO_API_BASE` env if needed.\r
- **Auth**: `Authorization: Bearer YOUR_API_KEY`\r
\r
### Request body (JSON)\r
\r
| Parameter | Type | Required | Default | Description |\r
|-----------|------|----------|---------|-------------|\r
| url | string | Yes | - | Webpage URL to extract |\r
| crawl_mode | string | No | fast | `fast` or `fine` |\r
| output_format | string | No | html | `html`, `text`, `markdown` |\r
| with_readability | boolean | No | - | Use readability (main content) |\r
| with_links_summary | boolean | No | - | Include links summary |\r
| with_images_summary | boolean | No | - | Include images summary |\r
| target_selector | string | No | - | CSS selector for target element |\r
| wait_for_selector | string | No | - | Wait for selector before extract |\r
| timeout | integer | No | - | Timeout in milliseconds |\r
| with_cache | boolean | No | true | Use cache |\r
\r
### Response\r
\r
Success (200):\r
\r
```json\r
{\r
  "code": 0,\r
  "message": "success",\r
  "data": {\r
    "content": { ... }\r
  }\r
}\r
```\r
\r
Extracted content is in `data.content`; structure depends on `output_format`.\r
\r
### Error codes\r
\r
| HTTP | Code | Description |\r
|------|------|-------------|\r
| 400 | - | Parameter validation failed |\r
| 401 | INVALID_API_KEY | API key invalid or revoked |\r
| 500/502 | WEB_EXTRACT_FAILED | Extract failed (server or page error) |\r
\r
## Output Format\r
\r
On success (script without `--json`):\r
\r
- Print the extracted content only (for direct use or piping).\r
\r
With `--json`:\r
\r
- Print full API response including `code`, `message`, `data`.\r
\r
Error response to user:\r
\r
```markdown\r
## Web Extract Failed\r
\r
- Error: \x3Ccode or message>\r
- URL: \x3Crequested url>\r
- Suggestion: \x3Ce.g. check URL, retry, or use --timeout>\r
```\r
\r
## Important Notes\r
\r
- Always check `FELO_API_KEY` before calling; if missing, return setup instructions.\r
- For long articles or slow sites, consider `--timeout` or `timeout` in request body.\r
- Use `output_format: "markdown"` and `with_readability: true` for clean article text.\r
- API may cache results; use `with_cache: false` in body only when fresh content is required (script does not expose this by default).\r
\r
## References\r
\r
- [Felo Web Extract API](https://openapi.felo.ai/docs/api-reference/v2/web-extract.html)\r
- [Felo Open Platform](https://openapi.felo.ai/docs/)\r