scraper-builder

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill frequently instructs the agent to execute shell commands using curl to interact with Bright Data APIs (e.g., querying the Dataset List API or triggering pre-built scrapers). It also references the execution of local shell scripts such as scripts/datasets.sh, skills/scrape/scripts/scrape.sh, and scripts/fetch.sh to perform scraping tasks.
  • [DYNAMIC_EXECUTION]: A core feature of the skill is generating custom Python or Node.js scraping scripts at runtime based on the target website's structure. The instructions explicitly mandate that the agent must run these generated scripts to validate extracted data before presenting it to the user.
  • [EXTERNAL_DOWNLOADS]: The skill utilizes several standard external libraries for its operations, including requests, beautifulsoup4, playwright, and aiohttp. It also fetches remote configuration and documentation from docs.brightdata.com and api.brightdata.com, which are official vendor domains.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from the open web (HTML, JSON, and metadata) during the reconnaissance and extraction phases.
  • Ingestion points: Target website content fetched via Web Unlocker or Browser API in SKILL.md and references/site-analysis-guide.md.
  • Boundary markers: Not explicitly defined in the provided code templates; the skill relies on structural parsing (CSS/XPath selectors) to isolate data.
  • Capability inventory: Access to shell commands, Python execution, and external network requests via Bright Data proxies.
  • Sanitization: The skill focuses on data extraction for display or storage; it does not provide specific sanitization for content that might be re-interpolated into future prompts.
  • [DATA_EXFILTRATION]: While the skill's purpose is data extraction, it operates within the expected scope of a scraping tool. It uses environment variables (BRIGHTDATA_API_KEY, BROWSER_AUTH) for authentication with vendor services, which is a standard and safe practice for this category of tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 12:44 PM
Security Audit — agent-trust-hub — scraper-builder