publish
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell commands including
git status --porcelainandgrepto inspect the project state and search for specific patterns in the codebase during troubleshooting. - [DYNAMIC_EXECUTION]: The skill invokes developer tools such as the TypeScript compiler (
pnpm exec tsc --noEmit) and the Next.js development server (pnpm dev) to validate application code and reproduce runtime issues. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data, specifically error logs and the project's source code, which serves as a potential vector for indirect prompt injection.
- Ingestion points: Source code files via
grepand runtime/build error output provided by the environment or tools (SKILL.md). - Boundary markers: No specific delimiters or instructions to ignore embedded content are used when processing logs or code.
- Capability inventory: The skill possesses deployment capabilities via
publish_site, source control modification viagit, and shell command execution for diagnostics. - Sanitization: The instructions do not specify any validation or sanitization for the content read from error logs or the codebase.
Audit Metadata