skills/olino3/forge/java-architect/Gen Agent Trust Hub

java-architect

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [Prompt Injection] (SAFE): No instructions attempting to override agent behavior or bypass safety filters were found.
  • [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials or sensitive local paths (such as SSH keys) are accessed. The skill only interacts with project-specific documentation.
  • [Unverifiable Dependencies & Remote Code Execution] (SAFE): No external code or scripts are included; the skill relies entirely on markdown-based instructions and built-in AI capabilities.
  • [Indirect Prompt Injection] (SAFE): The skill processes untrusted project documentation, which is a potential ingestion surface. However, the risk is negligible due to the lack of execution capabilities. Ingestion points: Files like architecture_decisions.md via memoryStore; Boundary markers: Absent; Capability inventory: No code execution, network access, or system-level writing; Sanitization: Absent.
  • [Persistence Mechanisms] (SAFE): No attempts to modify system startup files or schedule recurring tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:46 PM