ReasoningBank Intelligence
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (LOW): The skill requires the installation of external Node.js packages 'agentic-flow' and 'AgentDB'. These libraries are not maintained by the predefined list of trusted organizations (e.g., Google, Microsoft, Anthropic), which introduces a dependency risk if the packages are compromised or contain unvetted code.- [PROMPT_INJECTION] (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8) due to its core functionality of learning from task outcomes.
- Ingestion points: The methods 'recordExperience', 'learnPattern', and 'metaLearn' ingest data directly from agent tasks and environmental observations.
- Boundary markers: No specific delimiters or instructions to ignore embedded malicious content are present in the code snippets.
- Capability inventory: While the skill itself manages memory and strategy recommendations, it influences the agent's future 'strategies', which may include tool use or code execution.
- Sanitization: There is no evidence of validation or sanitization of 'outcomes' or 'observations' before they are incorporated into the long-term reasoning model.
Audit Metadata