agent-global-context-review

Pass

Audited by Gen Agent Trust Hub on May 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill follows its stated purpose of managing internal memory and lacks dangerous operations such as network access, remote code execution, or privilege escalation.
  • [PROMPT_INJECTION]: The skill reviews staged context, which is a potential surface for indirect injection.
  • Ingestion points: Memory candidates are read from staging files in the global context directory (~/.agent-global-context/).
  • Boundary markers: No explicit delimiters are mentioned for the content within candidate files.
  • Capability inventory: Limited to file operations within the designated memory directory.
  • Sanitization: Promotion rules exclude secrets and require user approval for inferred content.
Audit Metadata
Risk Level
SAFE
Analyzed
May 11, 2026, 12:34 PM