The Agent Skills Directory

[PROMPT_INJECTION]: The skill acts as a conduit for indirect prompt injection via its delegation workflow.\n
Ingestion points: The dispatch function in run.py accepts task and context parameters from the user or another agent without validation.\n
Boundary markers: The stored JSON task files do not use delimiters or instructions to prevent downstream agents from obeying commands embedded within the task text.\n
Capability inventory: The skill performs file writes to ~/.alex/tasks and provides command-line execution hints for downstream tools like alex exec.\n
Sanitization: There is no evidence of sanitization, escaping, or filtering of the input task content before it is persisted.

task-delegation