snyk-hardened

Pass

Audited by Gen Agent Trust Hub on Apr 21, 2026

Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [SAFE]: The skill implements a secure workflow that mandates authentication status checks and utilizes official Snyk MCP tools for scanning, demonstrating a well-defined and secure integration pattern.
  • [DATA_EXFILTRATION]: The skill's primary function involves scanning sensitive assets like source code and infrastructure configurations. This creates a data exposure surface where findings and metadata are processed within the AI context, though this is inherent to the tool's intended security purpose.
  • [PROMPT_INJECTION]: The skill uses extensive behavioral framing and self-referential safety claims (e.g., 'Armored by Faberlens'). These instructions attempt to define the skill's safety boundaries and persona within the prompt content.
  • [EXTERNAL_DOWNLOADS]: The skill and its safety documentation reference external technical reports and configuration pages on the vendor's official domain (faberlens.ai).
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 21, 2026, 06:31 PM
Security Audit — agent-trust-hub — snyk-hardened