skills/coleam00/archon/playwright-cli/Gen Agent Trust Hub

playwright-cli

Warn

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: MEDIUMDYNAMIC_EXECUTIONCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The skill makes extensive use of playwright-cli eval and playwright-cli run-code to execute arbitrary JavaScript strings within the browser context. It also supports loading and executing code from local files via the --filename argument, which constitutes dynamic loading of executable content.
  • Evidence:
  • playwright-cli eval "document.title" in SKILL.md
  • playwright-cli run-code "async page => { ... }" in references/running-code.md
  • playwright-cli run-code --filename=script.js in references/video-recording.md
  • [CREDENTIALS_UNSAFE]: The skill provides commands to extract, save, and load sensitive browser state, including session cookies and localStorage data, which often contain authentication tokens. The documentation explicitly suggests saving this state to local files (e.g., auth.json).
  • Evidence:
  • playwright-cli state-save auth.json in SKILL.md
  • playwright-cli cookie-get session_id in references/storage-state.md
  • playwright-cli localstorage-list in references/storage-state.md
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content from external, untrusted websites, creating a large attack surface for indirect prompt injection where malicious site content could influence the agent's logic.
  • Evidence:
  • Ingestion points: playwright-cli snapshot, playwright-cli find, and playwright-cli eval (when reading DOM content) in SKILL.md.
  • Boundary markers: None identified in the instructions to prevent the agent from obeying instructions embedded in web pages.
  • Capability inventory: The skill allows for file writing (screenshot, pdf, state-save), network interaction (via the browser), and arbitrary JavaScript execution.
  • Sanitization: No explicit sanitization of web content is mentioned before it is processed by the agent.
  • [EXTERNAL_DOWNLOADS]: The documentation encourages the global installation of the @playwright/cli package from the NPM registry.
  • Evidence:
  • npm install -g @playwright/cli@latest in SKILL.md
  • [COMMAND_EXECUTION]: The skill requires access to powerful shell commands including playwright-cli, npx, and npm to function, as defined in its allowed-tools configuration.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 15, 2026, 03:03 AM
Security Audit — agent-trust-hub — playwright-cli