huashu-prompt-save
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data by capturing arbitrary prompts and writing them to the local file system, creating a potential vector for indirect instructions.
- Ingestion points: The skill ingests user-provided text prompts through natural language interactions as described in SKILL.md.
- Boundary markers: The instructions lack explicit delimiters or instructions to ignore executable content or directives within the prompts being processed.
- Capability inventory: The skill performs file-write operations to the local directory to save markdown files and update an index file.
- Sanitization: There is no evidence of sanitization or filtering to prevent malicious payloads within the user-provided prompts from affecting downstream processing or agent behavior.
Audit Metadata