browserstack
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources. 1. Ingestion points: Build names, session metadata, and log content fetched via MCP tools browserstack_get_builds and browserstack_get_logs. 2. Boundary markers: None present in the instructions. 3. Capability inventory: Shell command execution via npx playwright test and file system writes to playwright.config.ts. 4. Sanitization: No explicit sanitization or filtering of external log content is described. This represents a standard data ingestion surface required for the skill's primary purpose.
- [EXTERNAL_DOWNLOADS]: The skill installs the browserstack-local package from the official npm registry, which is a standard dependency for BrowserStack testing.
- [DATA_EXFILTRATION]: The skill manages BROWSERSTACK_USERNAME and BROWSERSTACK_ACCESS_KEY environment variables to authenticate with the official BrowserStack service at cdp.browserstack.com. While credentials are included in connection parameters, this follows the official protocol for the well-known service.
- [COMMAND_EXECUTION]: Uses shell commands to execute Playwright tests and perform environment configuration, which are necessary for the skill's stated functionality.
Audit Metadata