functions

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFECREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: Instructs users to store sensitive API keys and project IDs in a local .env file for authentication.
  • [EXTERNAL_DOWNLOADS]: Downloads the official @browserbasehq/sdk-functions and playwright-core packages from the NPM registry to enable automation capabilities.
  • [COMMAND_EXECUTION]: Uses the vendor's bb CLI tool to initialize projects, start a local development server, and publish automation code to the cloud.
  • [DATA_EXFILTRATION]: Performs network requests to api.browserbase.com to manage function deployments and retrieve task results.
  • [PROMPT_INJECTION]: Vulnerable to indirect prompt injection via the browser automation engine which can load arbitrary external content.
  • Ingestion points: Ingests untrusted data when navigating the browser to URLs provided through input parameters in index.ts.
  • Boundary markers: The skill does not implement delimiters or instructions to ignore embedded commands within the content of navigated web pages.
  • Capability inventory: Supports comprehensive browser control (navigation, interaction, extraction) and communication with the platform's API.
  • Sanitization: The provided examples do not demonstrate validation or sanitization of input URLs or parameters before they are used in browser operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 02:03 AM