bright-data-best-practices

Pass

Audited by Gen Agent Trust Hub on Oct 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill provides an installation command for the Bright Data CLI that downloads a script from https://cli.brightdata.com/install.sh and executes it via the shell. This is a documented installation method for the vendor's own tooling.
  • [EXTERNAL_DOWNLOADS]: References the @brightdata/cli package for installation via NPM and the npx runner.
  • [INDIRECT_PROMPT_INJECTION]: The skill is intended to scrape external web content, which constitutes an attack surface for indirect prompt injection. The documentation specifically promotes a markdown output format optimized for consumption by large language models.
  • Ingestion points: External web data retrieved through various APIs or browser automation snippets in SKILL.md and the reference files.
  • Boundary markers: The provided code examples do not include explicit delimiters or instructions to ignore embedded prompts in the scraped data.
  • Capability inventory: The skill facilitates network operations and provides access to the bdata CLI, which possesses data retrieval and management capabilities.
  • Sanitization: There is no evidence of content sanitization or filtering in the provided integration examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 12, 2026, 11:25 PM