ReasoningBank Intelligence

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE] (SAFE): No malicious activity or security vulnerabilities were detected. The skill provides legitimate meta-cognitive features for AI agents using local storage and established internal packages.
  • [Indirect Prompt Injection] (LOW): The skill's learning mechanism creates a surface for indirect prompt injection through task outcomes. 1. Ingestion points: Experience data enters the system via 'rb.recordExperience' and 'rb.learnPattern' in SKILL.md. 2. Boundary markers: No delimiters or ignore instructions are present in the code snippets. 3. Capability inventory: Learned patterns influence 'rb.recommendStrategy' which guides agent behavior. 4. Sanitization: No input validation or filtering of experience data is shown in the integration examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:41 PM