to-spec
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from the conversation context and codebase understanding to generate its output.
- Ingestion points: Dialog history and repository files (SKILL.md).
- Boundary markers: None explicitly defined to distinguish instructions from user-provided data.
- Capability inventory: Filesystem exploration and network operations (posting to issue tracker).
- Sanitization: No explicit instructions for escaping or validating the content extracted from the context before interpolation.
- [COMMAND_EXECUTION]: The skill instructs the agent to publish the generated specification to an external issue tracker. This operation typically involves executing CLI commands or API calls (e.g., GitHub CLI) to perform network-based write operations.
Audit Metadata