chrome-devtools
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions require the execution of various shell commands, including npm install, npx browser installation, and a local install-deps.sh script to set up the environment.
- [PRIVILEGE_ESCALATION]: The documentation directs users to use sudo for installing system libraries and ImageMagick on Linux/WSL platforms, granting root access to the system package manager.
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external websites, creating an attack surface for indirect prompt injection where malicious web content could attempt to influence agent behavior.
- Ingestion points: Web pages navigated to via navigate.js or inspected with snapshot.js and evaluate.js (referenced in SKILL.md).
- Boundary markers: No specific delimiters or instructions to ignore embedded instructions are described in the documentation.
- Capability inventory: The skill can execute JavaScript in the browser (evaluate.js), capture screenshots, interact with forms, and monitor network traffic.
- Sanitization: The provided files do not mention any sanitization or filtering of the retrieved web content.
- [EXTERNAL_DOWNLOADS]: Installation involves downloading the Puppeteer library and the Chromium browser from official registries and Google's infrastructure.
- [DYNAMIC_EXECUTION]: The evaluate.js script allows for the dynamic execution of JavaScript code within the context of the automated browser session, which is a core feature of the skill.
Audit Metadata