ai-security-hardening

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill provides robust defensive patterns for detecting and neutralizing prompt injection attempts.
  • [EXTERNAL_DOWNLOADS]: Recommends the installation of reputable security tools like 'modelscan' and references libraries such as 'NeMo Guardrails' and 'Microsoft Presidio'.
  • [COMMAND_EXECUTION]: Includes standard shell operations for verifying model weights using SHA-256 hashing to prevent tampering.
  • [PROMPT_INJECTION]: Demonstrates an API surface for handling untrusted data via a FastAPI endpoint.
  • Ingestion points: User-provided text enters the system through the chat completion endpoint in SKILL.md.
  • Boundary markers: Implements a dedicated input validation and sanitization layer.
  • Capability inventory: Processes inputs for LLM generation and performs automated output safety validation.
  • Sanitization: Employs regex filters to strip control characters and block malicious injection strings.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 11:44 PM