memory-hierarchy

Pass

Audited by Gen Agent Trust Hub on Mar 22, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [NO_CODE]: The skill is entirely informational and does not contain any functional code, scripts, or tool definitions. It is safe to load as it performs no actions on its own.
  • [SAFE]: The technical descriptions of file paths (e.g., ~/.claude/CLAUDE.md) and configuration hierarchies are consistent with the documented behavior of the Claude Code platform. There is no evidence of data exfiltration or credential theft attempts.
  • [PROMPT_INJECTION]: The examples provided in the documentation are benign coding standards and workflow preferences. They do not attempt to manipulate the agent's safety protocols or override system instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 22, 2026, 10:32 PM
Security Audit — agent-trust-hub — memory-hierarchy