dev-desktop-sandbox
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The workflow documentation involves running local shell commands such as
make,bunx electron, and build scripts to initialize and launch a sandboxed Electron environment. - [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface through the ingestion of local configuration files. 1. Ingestion points:
providers.jsoncandconfig.jsonin the root directory. 2. Boundary markers: None explicitly defined in the workflow. 3. Capability inventory: Subprocess execution viamakeandbunx. 4. Sanitization: None described. - [SAFE]: The documentation demonstrates security-conscious practices by isolating instance data in temporary directories and explicitly recommending the exclusion of sensitive files like
secrets.jsonfrom the sandbox state.
Audit Metadata