The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill accepts user-provided input through the $ARGUMENTS variable and interpolates it into a structured briefing for a new sub-agent created via the create_agent tool.
Ingestion points: Untrusted data enters the agent context through the $ARGUMENTS placeholder in SKILL.md.
Boundary markers: The skill employs a structured Markdown template (headers such as ## Task, ## Context, ## Acceptance criteria) to delimit the handoff information.
Capability inventory: The skill uses create_agent, create_workspace, and list_profiles to manage the transition and environment.
Sanitization: The instructions do not specify explicit sanitization or filtering of the user input, relying on the structural template to maintain context.

paseo-handoff