graphiti
Pass
Audited by Gen Agent Trust Hub on Aug 7, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs the
graphiti-corelibrary from the official Python package registry. This package is maintained by Zep, a well-known provider of agentic memory solutions, and targets their official GitHub repository. - [PROMPT_INJECTION]: The skill is designed to process untrusted data (episodes) for ingestion into a knowledge graph. While the skill configuration allows powerful tools like
Bashand file editing, the potential for indirect prompt injection is a functional characteristic of memory-processing skills and is mitigated by the intended architectural use case.
Audit Metadata