skills/kensaurus/skills/test-red-team/Gen Agent Trust Hub

test-red-team

Pass

Audited by Gen Agent Trust Hub on Oct 7, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill installs and executes the Playwright CLI using npx --yes @playwright/cli@latest. Playwright is a well-known automation tool maintained by Microsoft, a trusted organization.
  • [EXTERNAL_DOWNLOADS]: The skill downloads the Playwright CLI package from the public npm registry at runtime.
  • [COMMAND_EXECUTION]: The skill utilizes playwright-cli for browser interaction and adb shell commands for mobile application testing and interaction.
  • [DYNAMIC_EXECUTION]: The skill uses eval and the Chrome DevTools Protocol (CDP) via Playwright to simulate specific environment conditions such as network latency and throttling for performance testing.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted application data and source code as part of its auditing function, which creates an indirect prompt injection attack surface. Ingestion points: The agent reviews UI screenshots, network responses, database records, and project source files. Boundary markers: Instructions specify using the RT-TEST- prefix for all testing data to ensure isolation from production data. Capability inventory: The skill provides access to browser automation (Playwright), Android shell debugging (ADB), and database interaction (supabase:execute_sql). Sanitization: The instructions mandate using only benign payloads for vulnerability verification and strictly prohibit the storage of exploit proof-of-concepts within the repository.
  • [DATA_EXFILTRATION]: The skill accesses sensitive session information, authentication profiles, and database records to test security boundaries (e.g., IDOR flaws). This is within the scope of its primary security auditing function and no unauthorized external exfiltration was identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 7, 2026, 11:27 PM