The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted lead data fetched from HubSpot lists, which creates a surface for indirect prompt injection attacks where malicious data in lead fields could influence the agent's routing or qualification logic. • Ingestion points: Lead data is ingested via hubspot_list_members and hubspot_find_contacts_by_role tools as described in Stage 1 and Stage 6 of SKILL.md. • Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings for external data. • Capability inventory: The skill possesses the capability to update CRM records via hubspot_update_contact, manage enrollment states via createSequenceState and finishSequenceState, and create tasks via hubspot_create_contact_task. • Sanitization: There is no mention of sanitizing, escaping, or filtering external content before it is processed by the qualification and routing logic.
[COMMAND_EXECUTION]: The skill executes multiple shell-like commands through its integrated MCP tools (Epiphan AI and Nooks) to perform actions like lead enrichment, list searching, and sequence management.

nooks-autopilot