conversation-memory

Pass

Audited by Gen Agent Trust Hub on Feb 15, 2026

Risk Level: LOWNO_CODE
Full Analysis
  • [No Code] (SAFE): The provided file contains only Markdown documentation and metadata. There are no shell scripts, Python code, or JavaScript files associated with this skill.
  • [Data Exposure] (INFO): The documentation explicitly acknowledges the risk of cross-user data leakage in its 'Sharp Edges' section, recommending strict user isolation as a solution.
  • [Indirect Prompt Injection] (LOW): The skill's purpose (retrieving and inserting memories into prompts) creates an inherent surface for indirect prompt injection, but no exploitable implementation is present in the source file provided.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 15, 2026, 08:50 PM