verify-ui
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and install a browser-based devkit from an external GitHub repository.
- Evidence: Refers to
https://github.com/DevSports/devkitfor the browser tool and usespnpm browser:installto set up the environment. - [COMMAND_EXECUTION]: The skill uses the shell to execute project-specific commands and browser automation tasks.
- Evidence: Frequent use of
pnpmfor authentication (pnpm auth), browser control (pnpm browser), and tool installation. - [INDIRECT_PROMPT_INJECTION]: The skill performs automated browsing and analysis of web page content, which is a common vector for indirect prompt injection.
- Ingestion points: Headless Chrome browser snapshots, console logs, and network request data (
SKILL.mdsections 2 and 3). - Boundary markers: No explicit boundary markers or instructions to ignore embedded content within the analyzed UI are provided.
- Capability inventory: The agent can execute shell commands (
pnpm), navigate to arbitrary URLs, capture screenshots, and read page metadata. - Sanitization: No sanitization or validation of the ingested DOM content or console output is mentioned before the agent processes it for verification.
Audit Metadata