reality-verification
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill maps specific keywords to shell commands like pnpm test, pnpm build, eslint, and GitHub CLI operations (gh run view, gh api). These commands are executed locally to diagnose and verify the state of the project.
- [EXTERNAL_DOWNLOADS]: Utilizes MCP tools such as fetch and playwright to perform network requests and browser-based testing. This allows the agent to verify remote API endpoints and UI states.
- [PROMPT_INJECTION]: The skill processes untrusted data which could potentially contain malicious instructions.
- Ingestion points: Command output from test runners, GitHub action logs, and API responses.
- Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within the captured output.
- Capability inventory: Local shell execution, GitHub CLI access, and network/browser interaction via MCP tools.
- Sanitization: There is no evidence of sanitization or validation of the content captured from command outputs or network responses.
Audit Metadata