scrapling

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The scripts/install.sh script installs the scrapling package from PyPI and invokes scrapling install to download browser dependencies such as Playwright and Patchright. These downloads originate from official registries and are required for the framework's browser-based scraping capabilities.
  • [COMMAND_EXECUTION]: The skill uses local shell scripts (scripts/run-extract.sh, scripts/run-mcp.sh) to wrap the scrapling CLI. These scripts are used for legitimate tasks such as fetching website content and hosting a Model Context Protocol (MCP) server for agent integration.
  • [PROMPT_INJECTION]: The skill's primary function is to ingest data from external websites, which creates a surface for indirect prompt injection. Malicious content on a scraped page could contain instructions designed to influence the agent's subsequent actions.
  • Ingestion points: Data fetched from external URLs via Scrapling's fetcher classes.
  • Boundary markers: Not explicitly defined in the main instructions.
  • Capability inventory: Subprocess execution for the scraper CLI and file writing for output storage.
  • Sanitization: The skill converts HTML to Markdown or text, which provides a layer of structural isolation, but does not specify further content sanitization.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 02:29 AM