The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by interpolating untrusted user input {{args}} directly into the instructions for tool/skill invocation.
Ingestion Point: SKILL.md contains multiple instances where {{args}} is used.
Boundary Markers: Absent. There are no delimiters (e.g., XML tags or triple quotes) used to separate the user data from the agent's logic.
Capability Inventory: The skill is designed to invoke other functional skills such as surge-activation and surge-recon, effectively acting as a command dispatcher.
Sanitization: Absent. The instructions do not tell the agent to ignore or sanitize potentially malicious instructions embedded within the user input.

surge