playwriter
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_CONTEXT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: Executes arbitrary JavaScript snippets using the
playwriter -ecommand to facilitate browser automation and web interaction. - [EXTERNAL_DOWNLOADS]: Fetches the latest version of the
playwritertool from the official NPM registry vianpxorbunx. - [INDIRECT_PROMPT_INJECTION]: Processes untrusted content from external websites which could contain malicious instructions.
- Ingestion points: External web content retrieved via the browser through navigation or scraping commands.
- Boundary markers: No specific delimiters or safety warnings are provided in the skill manifest.
- Capability inventory: JavaScript execution, local file writing (screenshots), and network access.
- Sanitization: No explicit sanitization or filtering of ingested web content is described in the documentation.
- [DYNAMIC_CONTEXT_INJECTION]: Directs the agent to fetch the full instructional set at runtime using the
playwriter skillcommand, which allows for dynamic modification of the agent's behavior.
Audit Metadata