to-prd
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by ingesting and summarizing external, potentially untrusted data from the codebase and conversation context.
- Ingestion points: The skill is instructed to take "the current conversation context and codebase understanding" as input for the PRD generation (found in SKILL.md).
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish between its primary instructions and potentially malicious instructions embedded in the codebase or conversation data.
- Capability inventory: The skill has the capability to write and "publish it [the PRD] to the project issue tracker," which constitutes a data-writing sink.
- Sanitization: The instructions do not include steps to sanitize, escape, or filter content extracted from the repository or conversation before publishing it to the issue tracker.
Audit Metadata