daytona-electron-den
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell scripts located in the local
.devcontainerdirectory (test-server-on-daytona.shandtest-on-daytona.sh) to initialize testing environments. It also utilizes thedaytona execcommand to read configuration files (desktop-bootstrap.json) and run commands within remote sandboxes. - [INDIRECT_PROMPT_INJECTION]: The skill provides an attack surface by ingesting untrusted data from the application's renderer and environment variables.
- Ingestion points: Reads text from the Electron DOM (
document.body.innerText) and uses environment variables (DEN_WEB_URL,DEN_API_URL) to construct deep links and API requests. - Boundary markers: The instructions do not define explicit boundary markers or delimiters for the ingested data.
- Capability inventory: The skill has the capability to execute shell scripts, run remote commands via
daytona exec, and perform network operations viacurl. - Sanitization: There is no evidence of sanitization or validation performed on the captured DOM text before processing.
Audit Metadata