rpg-persona
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from user-provided input and project documents such as
docs/CLIENT_PERSONAS.mdanddocs/ICP_ANALYSIS.md. This ingestion surface could potentially be used for indirect prompt injection if the files contain malicious instructions. - Ingestion points: User chat messages and local markdown files (
docs/CLIENT_PERSONAS.md,docs/ICP_ANALYSIS.md). - Boundary markers: The skill uses a strict markdown structure for responses (alternating Buyer and Coach blocks) which helps maintain role separation.
- Capability inventory: The skill is limited to reading local files and writing session notes to
docs/ROLEPLAY_NOTES.md. It has no network exfiltration or command execution capabilities. - Sanitization: The instructions include a specific rule (Rule 13) to ask the user before including confidential or non-public data in the saved artifacts, providing a manual oversight step.
Audit Metadata