design-procedural-generation
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill demonstrates a strong security posture by including a dedicated 'Security Guardrails' section. It explicitly warns against evaluating content-authored constraints as code (ASI05) and mandates the use of schema-validated data files. This proactive approach significantly reduces the risk of remote code execution.\n- [INDIRECT_PROMPT_INJECTION]: The skill manages an attack surface where external data (master seeds, tile constraints, and generator configurations) is processed to drive generation logic.\n
- Ingestion points: Reads master seeds from save files or logs and loads WFC tile constraints or generator configurations from external data files (SKILL.md).\n
- Boundary markers: The instructions mandate validating seed format, length, and ranges, and using schema-validated data files to delimit external content.\n
- Capability inventory: The skill has access to tools including
execute_command,write_file, andrun_build(SKILL.md).\n - Sanitization: It explicitly requires rejecting executable payloads or scripts within tile packs and sanitizing untrusted seeds before derivation.
Audit Metadata