The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill grounds agent responses in external markdown files, creating a potential surface for instructions embedded in those files to influence agent behavior.\n- Ingestion points: references/patterns.md, references/sharp_edges.md, and references/validations.md.\n- Boundary markers: Absent. The agent is instructed to treat these files as the source of truth without warnings to ignore embedded directives.\n- Capability inventory: No dangerous tools, executable scripts, subprocess calls, or network operations are included in the skill.\n- Sanitization: No validation or sanitization of the referenced content is performed.

disaster-recovery