ReasoningBank with AgentDB
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides documentation for ReasoningBank and AgentDB, which are legitimate tools for agent learning and memory management.
- [EXTERNAL_DOWNLOADS]: The skill utilizes
npx agentdb@latest, a standard method for running the vendor's command-line tools. - [COMMAND_EXECUTION]: The documentation includes CLI commands for database initialization and MCP server setup, which are core functions of the described library.
- [INDIRECT_PROMPT_INJECTION]: The memory system architecture stores agent trajectories and user experiences, creating a surface for indirect prompt injection. 1. Ingestion points:
rb.insertPatterninSKILL.md. 2. Boundary markers: Absent in snippets. 3. Capability inventory: Data retrieval and synthesis via API and CLI. 4. Sanitization: Not explicitly shown. This represents a known design characteristic of learning-based agent systems.
Audit Metadata