The Agent Skills Directory

Prompt Injection (LOW): The skill instructs the agent to ground all responses in external files (references/patterns.md, references/sharp_edges.md, references/validations.md) and to 'ignore generic approaches' in favor of these patterns, establishing a surface for indirect prompt injection. 1. Ingestion points: references/patterns.md, references/sharp_edges.md, and references/validations.md. 2. Boundary markers: Absent; there are no instructions to disregard potentially malicious commands embedded in the reference data. 3. Capability inventory: None; the skill is purely instructional and lacks scripts, subprocess calls, or network access. 4. Sanitization: Absent; no validation or escaping logic is defined for the external data.

animation-systems