Extract structured data, clean reader-mode markdown, search results, and batch-fill forms via agents-browser MCP without writing custom ad-hoc scraping scripts. Runs headless by default.
Installs into .claude/skills of the current project.
Are you the author of Browser Scrape?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/lolaplex-browser-scrape)
---
name: browser-scrape
description: Extract structured data, clean reader-mode markdown, search results, and batch-fill forms via agents-browser MCP without writing custom ad-hoc scraping scripts. Runs headless by default.
---
# Browser Scrape & Search Skill
Use this skill for web information retrieval, structured table extraction, search queries, and form automation.
**Headless is default.** Do not call `visible=True` or `browser_system_open` unless the human explicitly asks for a desktop window.
## High-Level Operations
1. **Direct Web Search**:
- `browser_search(query="python packaging guide")`
- Returns top organic results with Title, URL, and Snippet in one call.
2. **Clean Article Reader Mode**:
- `browser_read_article()`
- Removes navbars, headers, footers, ads, and sidebars. Returns clean markdown article content.
3. **Structured Selector Scraping**:
- `browser_scrape(selector="table.stats", mode="table")` $\rightarrow$ Markdown table
- `browser_scrape(selector=".doc-links", mode="links")` $\rightarrow$ Link list
- `browser_scrape(selector="main", mode="text")` $\rightarrow$ Clean text
4. **Batch Form Filling**:
- `browser_fill_form(fields_json='{"@1": "admin", "@2": "password123"}', submit_selector="@3")`
