memory-system
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill establishes a persistent memory system by reading and writing files in the
.agents/memory/directory. This creates a surface for indirect prompt injection where malicious instructions saved in one session could be loaded and followed by the agent in a future session.\n - Ingestion points: The agent reads the
.agents/memory/MEMORY.mdindex and associated topic files (e.g.,user-preferences.md) during the Recall and Search operations defined inSKILL.md.\n - Capability inventory: The skill utilizes
Read,Write,Grep, andGlobtools to manage and search the memory directory.\n - Boundary markers: The instructions do not specify the use of delimiters or protective framing (e.g., XML tags or 'ignore instructions' wrappers) when loading memory content into the prompt context.\n
- Sanitization: There is no defined process for sanitizing or validating memory entries to ensure they contain data only and not executable instructions.
Audit Metadata