openviking-memory

Pass

Audited by Gen Agent Trust Hub on Oct 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the retrieval of long-term semantic memory and the ingestion of external resources (URLs/documents), which creates a surface for potential indirect prompt injection from stored or external content.
  • Ingestion points: External and prior data enters the agent's context via the find, search, read, and add_resource tools.
  • Boundary markers: The instructions do not define technical delimiters for retrieved memory content, but they include strong behavioral constraints, directing the agent to prioritize system instructions and current user requests over retrieved memory.
  • Capability inventory: The skill includes tools for persisting memories (remember), importing external resources (add_resource), and optionally modifying files (write, edit) within the memory scope.
  • Sanitization: The skill explicitly warns against persisting secrets or credentials and mandates that the agent verify all retrieved commands, paths, and versions against the current task environment before execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 29, 2026, 08:35 AM