synapse
Pass
Audited by Gen Agent Trust Hub on Mar 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a comprehensive and structured context engine for managing agent behavior. It uses localized file paths (e.g., .synapse/, .aiox-core/) consistent with the vendor's documented architecture and provides clear documentation for all operational commands.\n- [PROMPT_INJECTION]: The skill's primary function is the intentional injection of context-aware rules into user prompts via the UserPromptSubmit hook. This is a core architectural feature rather than a safety bypass, and the system uses XML markers to maintain instructional hierarchy and prevent unintended instruction confusion.\n- [COMMAND_EXECUTION]: The skill provides several CRUD commands (e.g., *synapse create, *synapse edit) that modify local configuration files. These commands are part of the intended management interface and do not expose the system to arbitrary shell execution or unauthorized external access.
Audit Metadata