The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which could potentially influence agent behavior or cause the execution of unintended tool calls.
Ingestion points: The skill ingests data from external sources using tools like list_client_notes, get_client_note, list_incidents, list_care_plans, and list_quotes (SKILL.md).
Boundary markers: The instructions do not specify the use of boundary markers or delimiters when processing these external data fields to separate data from instructions.
Capability inventory: The skill utilizes tools with significant write capabilities, including create_shift, create_invoice_payment, create_progress_note, update_shift, and lifecycle management tools like enable_care_plan (SKILL.md).
Sanitization: The skill provides robust mitigation by mandating that the agent must "Obtain the user's explicit confirmation immediately before every write tool call" and "verify with the corresponding read tool" after any mutation.

shiftcare-basics