demo
Warn
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [DYNAMIC_EXECUTION]: The skill instructs the agent to generate and write JavaScript code (Playwright gestures) into a file (play.mjs) that is later executed by the environment.
- [DYNAMIC_EXECUTION]: The automation script play.mjs dynamically loads the Playwright module using a computed path that includes the user's home directory.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local repository (README, routes, dashboard files) and the application's UI to populate the demo deck's content.
- Ingestion points: Local repository files and application UI elements via Playwright.
- Boundary markers: The agent is instructed to request manual confirmation from the participant for all generated content.
- Capability inventory: The skill possesses file system write permissions and the ability to execute shell commands.
- Sanitization: The render.mjs script applies HTML entity escaping to text interpolated into the final demo deck.
- [COMMAND_EXECUTION]: The skill executes various command-line tools including npm, npx, and vendor-provided CLI utilities for setup, execution, and deployment.
- [EXTERNAL_DOWNLOADS]: The skill installs the Playwright library and vendor-owned CLI tools from the public NPM registry.
Audit Metadata