brightdata-cli

Fail

Audited by Gen Agent Trust Hub on Oct 29, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The installation instructions for macOS and Linux recommend downloading a shell script from the vendor's domain (https://cli.brightdata.com/install.sh) and piping it directly to bash for execution. This resource is provided by the skill author.
  • [DATA_EXFILTRATION]: The documentation explicitly identifies the storage locations for sensitive authentication credentials (API keys) at ~/.config/brightdata-cli/credentials.json (Linux/macOS) and %APPDATA%\brightdata-cli\credentials.json (Windows), exposing them to potential agent-led exfiltration if prompted.
  • [INDIRECT_PROMPT_INJECTION]: The skill enables the ingestion of untrusted external content via web scraping (bdata scrape) and search engine results (bdata search), creating an attack surface for instructions embedded in third-party data.
  • Ingestion points: Content retrieved from external URLs and search engine organic results.
  • Boundary markers: The instructions do not define usage of delimiters or safety warnings to distinguish between scraped data and agent instructions.
  • Capability inventory: The CLI tool possesses capabilities to write to the local filesystem and perform authenticated network requests.
  • Sanitization: No explicit sanitization or filtering of retrieved web content is described before it enters the agent's context.
  • [COMMAND_EXECUTION]: The skill provides numerous examples for executing shell commands with the brightdata binary, including the use of piping and redirection which could be manipulated for unintended system operations.
Recommendations
  • HIGH: Downloads and executes remote code from: https://cli.brightdata.com/install.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Oct 29, 2026, 04:30 PM