prove-it
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands for running project tests and UI automation. Specifically, it mentions
npm testandnpx playwright install chromiumto verify application behavior and capture screenshots. - [EXTERNAL_DOWNLOADS]: The skill references the installation of the Chromium browser via Playwright (
npx playwright install chromium). This is a standard procedure for UI testing using well-known automation frameworks. - [INDIRECT_PROMPT_INJECTION]: The skill involves processing user feedback about recurring mistakes and reading from project-specific files like
AGENTS.md. This represents a potential attack surface where external input could influence the agent's logic, although the instructions focus on implementing defensive code and documentation. - [DYNAMIC_EXECUTION]: The agent is prompted to generate and execute new test scripts (e.g., unit tests or Playwright browser tests) to validate code changes. While this involves generating executable code, it is a primary function of the developer-oriented skill.
Audit Metadata