playwright-stealth-verify

Pass

Audited by Gen Agent Trust Hub on Aug 5, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Installs the 'liarjs' package from the NPM registry and executes it using 'npx'. These are vendor-owned resources consistent with the skill's purpose.
  • [DATA_EXFILTRATION]: Makes a documented network request to 'https://liarjs.dev/api/net.json' to analyze how the browser appears to a remote server. This feature can be opted out of using the '--offline' flag.
  • [COMMAND_EXECUTION]: Uses system commands like 'npm' and 'npx' to set up and run the audit tool.
  • [PROMPT_INJECTION]: The skill uses 'page.evaluate()' to inspect browser state, which allows the agent to ingest data from web pages. While the skill instructs the agent not to navigate to third-party sites, the interaction with external page content is an ingestion point for potential indirect prompt injection. 1. Ingestion points: Browser page content accessed via 'page.evaluate()' in SKILL.md. 2. Boundary markers: Includes an explicit warning to avoid third-party sites during scans. 3. Capability inventory: Access to Bash, Read, Edit, and Write tools as specified in the skill frontmatter. 4. Sanitization: No specific sanitization or filtering of the browser's response data is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 5, 2026, 01:56 AM
Security Audit — agent-trust-hub — playwright-stealth-verify