The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the internal_note field of prospect records retrieved via the classic_get_prospect tool. This creates a surface for indirect prompt injection attacks where instructions embedded in notes could influence the agent's behavior during triage or update operations.
Ingestion points: The classic_get_prospect tool returns internal_note and signup metadata used for analysis (referenced in SKILL.md).
Boundary markers: Absent. The skill instructions do not define delimiters or provide guidance to ignore instructions embedded within the prospect notes.
Capability inventory: The skill has write access via classic_update_prospect_note and suggests integration with other tools like membership-push-broadcast and membership-internal-subscription-grant.
Sanitization: Absent. There is no specified logic for escaping, filtering, or validating the content of the notes before processing.

membership-prospect-followup