what-would-a-senior-engineer-do
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted repository data such as source code, configuration files, and documentation. This creates a surface where malicious instructions hidden in those files could influence the agent's reasoning or actions.
- Ingestion points: Repository source code, tests, configuration, history, and documentation are read during the 'Core method' phase.
- Boundary markers: The skill does not define explicit delimiters to isolate untrusted repository data from the agent's internal instruction context.
- Capability inventory: The skill has the authority to read files, search repository history, modify code or configuration, and execute local build/validation tools.
- Sanitization: No instructions are provided to sanitize or validate the content ingested from the repository.
- [COMMAND_EXECUTION]: The instructions direct the agent to execute repository-specific tools, including type checks, tests, linters, and build commands, to verify implementation requests. This involves executing scripts or binaries present in the user's local development environment.
Audit Metadata