agent-safla-neural
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides architectural guidelines and instructional content for a specialized AI persona. The included code blocks are illustrative examples of using Model Context Protocol (MCP) tools for neural training and memory persistence.
- [INDIRECT_PROMPT_INJECTION]: The skill architecture explicitly involves storing and retrieving interaction histories and context in a persistent memory tier. While this creates a surface for indirect prompt injection should the agent process malicious data, this behavior is central to the skill's primary purpose of cross-session memory and context persistence. The implementation uses discrete tool calls for storage, which allows for platform-level management and sanitization.
Audit Metadata