character-asset

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill is a comprehensive prompt engineering guide and asset management protocol. It does not execute shell commands, access sensitive file system paths, or initiate network connections.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface as it processes user-provided character dimensions and facial features to generate prompts (Section 1, Section 10). Evidence Chain: (1) Ingestion points: User input for character dimensions in SKILL.md. (2) Boundary markers: Present as structured prompt templates and JSON interfaces. (3) Capability inventory: No high-risk capabilities (subprocess calls, network ops, or file writes) are present. (4) Sanitization: The skill includes a manual Quality Check list in Section 8 to verify that generated outputs adhere to defined constraints. The risk is assessed as safe given the lack of exploitable tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 02:23 AM
Security Audit — agent-trust-hub — character-asset