skills.rest
ruvnet
ruvnetCOMMUNITY
Explore all skills by this author →
Website𝕏@ruvGitHub

browser-extract

Extract structured data from web pages using reusable templates with PII and prompt-injection screening.

70.1k|8.4k|Updated Jun 2, 2025
Search Tags:#prompt injection#Templates#Browser Automation#pii redaction#dom parsing#Data extraction#Web scraping
One-click install
npx skills add https://github.com/ruvnet/claude-flow --skill browser-extract
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: browser-extract
Source: https://github.com/ruvnet/claude-flow/tree/main/plugins/ruflo-browser/skills/browser-extract
Command: npx skills add https://github.com/ruvnet/claude-flow --skill browser-extract

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Scraping rendered web pages often produces unstructured text that may contain personal data or hidden prompt-injection payloads, and one-off scrapes cannot be reused. This Skill extracts structured data from web pages while enforcing AIDefence PII redaction and prompt-injection checks before any content reaches the model or storage.

Core Features & Use Cases

  • Template-Based Extraction: Save selector chains as host-scoped browser-templates in AgentDB and re-run them against new URLs on the same host.
  • Recorded Sessions: Every extraction runs inside a recorded RVF session via browser-record, making runs replayable and auditable.
  • Mandatory Safety Gates: Every extracted string passes aidefence_has_pii for redaction and aidefence_is_safe for prompt-injection screening before returning to the model.
  • Use Case: Extract article titles and body text from a news site, save the selector chain as a template, and re-run it daily against new article URLs with PII automatically redacted.

Quick Start

Use the browser-extract skill to pull the article text from this URL and save the extraction pattern as a template named news_article.

Frequently Asked Questions about browser-extract

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract structured data from a web page with browser-extract?▼

Open a recorded session via browser-record, wait for dynamic content with browser_wait, then use browser_snapshot for accessibility trees or browser_eval with querySelectorAll for bulk lookups. Extracted strings pass PII and injection gates before being returned.

How do I save and reuse a scraping template?▼

Pass --save-template <name> during extraction to persist the selector chain to the browser-templates namespace in AgentDB. Later runs use --template <name> to retrieve and replay the recipe against new URLs on the same host.

Are extraction templates portable across different websites?▼

No, templates are host-scoped. A template built for one domain, such as theguardian.com, is not portable to another site like nytimes.com without re-validation of its selector chain.

What happens if the AIDefence MCP tools are not initialized?▼

The skill refuses to run and surfaces a doctor remediation instead. The AIDefence PII and prompt-injection gates are mandatory and cannot be bypassed for any extraction.

How does the skill handle paginated extractions?▼

The pagination cursor is persisted between pages in the trajectory step args, so the recorded trace alone is sufficient to replay the full multi-page extraction.

Related Skills

Explore More Software Engineering Skills→
redbananastudiosredbananastudios

browser-use

Automate browser interactions for web testing and data extraction via CLI.

Community
Intermediate
TurnaboutHeroTurnaboutHero
12

data-wizard

Automate ETL and data transformation for CSV, JSON, and database sources using Python and pandas.

Community
Advanced
ricardorochericardoroche
2

ai-security

Detect prompt injection attempts in AI and LLM applications.

Community
Intermediate
Z43LZ43L

scrapling

Extract web content from static, dynamic, and Cloudflare-protected pages.

Community
Advanced