character-portrait

Pass

Audited by Gen Agent Trust Hub on Nov 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security threats were detected during the analysis. The skill is instructional and focuses on scene planning for 3D character portraits.
  • [PROMPT_INJECTION]: No evidence of prompt injection or system instruction bypass attempts. The instructional language is focused on domain-specific constraints (e.g., verifying likeness rights).
  • [DATA_EXPOSURE]: No hardcoded credentials, sensitive file paths, or data exfiltration logic found. The skill does not perform any network operations.
  • [REMOTE_CODE_EXECUTION]: No external package dependencies or remote code execution patterns were identified. The skill contract (contract.json) confirms an L0 capability tier with no tool access.
  • [OBFUSCATION]: No obfuscated text, hidden URLs, or suspicious encoded strings (Base64, Hex, etc.) were found in the skill body or reference files.
  • [DYNAMIC_CONTEXT_INJECTION]: No use of dynamic command execution markers (e.g., backtick commands) in the SKILL.md file.
  • [INDIRECT_PROMPT_INJECTION]: Although the skill ingests external data (a brief from intent-to-3d-brief-writer), it lacks any executable capabilities or tools that could be exploited via malicious input. It strictly adheres to a planning-only boundary.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 7, 2026, 09:09 PM