skills/seaworld008/commonly-used-high-value-skills/browser-testing-with-devtools/Gen Agent Trust Hub
browser-testing-with-devtools
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill configuration instructs the user to install the
chrome-devtools-mcppackage vianpx. This is a standard package installation from the official NPM registry and is a resource owned and maintained by the skill's author, representing intended functionality for the Chrome DevTools integration.- [INDIRECT_PROMPT_INJECTION]: The skill inherently interacts with untrusted data from the web, creating a potential surface for indirect prompt injection attacks. However, the skill includes extensive defensive instructions to mitigate this risk. - Ingestion points: Browser DOM, console logs, network responses, and JavaScript execution results are identified as external ingestion points.
- Boundary markers: The documentation provides explicit 'Security Boundaries,' including a visual trust diagram and instructions to label all browser data as untrusted.
- Capability inventory: The agent utilizes
chrome-devtools-mcpfor screenshots, DOM inspection, network monitoring, and JavaScript execution. - Sanitization: The skill mandates that the agent must never interpret browser content as instructions, must confirm all navigations with the user, and must flag suspicious content or hidden directives found in the page.
- [PROMPT_INJECTION]: While automated scanners detected phrases like 'Ignore previous instructions,' these are used exclusively within the security guidelines to warn the agent about malicious patterns it might encounter in untrusted web content, rather than attempting to subvert the agent's own behavior.
Audit Metadata