memory-dream
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from daily chat logs and diary files to update long-term memory. 1. Ingestion points: Reads markdown files within the memory/ directory containing conversation history. 2. Boundary markers: The instructions do not specify explicit delimiters for the ingested logs. 3. Capability inventory: Uses file listing and editing tools on local markdown files. 4. Sanitization: No explicit content filtering is performed.
- [SAFE]: The skill includes safety rules against exposing credentials and uses precise editing tools to prevent data loss.
Audit Metadata