cherry-electron-dev
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell commands (
ps,lsof,kill,pnpm,git,gh) to discover, launch, and control Electron processes and manage source code branches. It includes specific procedures to ensure these commands target only the intended workspace and PIDs. - [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection by reading untrusted data from GitHub Pull Request metadata and file diffs. A malicious actor could embed instructions in a PR description or code comments intended to manipulate the agent's behavior during the review or testing process.
- Ingestion points: PR data is fetched in
references/pr-testing.mdusing theghCLI. - Boundary markers: No specific boundary markers are enforced for the PR body content.
- Capability inventory: The skill allows filesystem writes, shell execution, and JavaScript execution within the target app.
- Sanitization: The instructions do not mandate sanitization of the PR-sourced text.
- [DYNAMIC_EXECUTION]: The skill leverages the Chrome DevTools Protocol (CDP) to execute JavaScript code dynamically within the renderer and main processes of the Electron application. This is essential for the intended performance analysis and automated UI testing features.
Audit Metadata