The Agent Skills Directory

SAFE (SAFE): The skill consists of technical documentation and design guidelines for Unreal Engine audio systems. No evidence of prompt injection, data exfiltration, obfuscation, or unauthorized command execution was found.
Indirect Prompt Injection (SAFE): While the skill accesses external data templates, it specifies a robust 7-stage validation process to ensure graph integrity. Evidence Chain: 1. Ingestion points: src/ue_audio_mcp/templates/ via Read and Glob tools. 2. Boundary markers: Not explicitly defined in documentation. 3. Capability inventory: Asset creation and graph manipulation via a specialized Builder API. 4. Sanitization: Claimed 7-stage validation of required fields, interfaces, node types, and pin compatibility.

metasound-dsp