The Agent Skills Directory

Indirect Prompt Injection Surface: The skill processes untrusted user input from the $ARGUMENTS variable in Step 1 to extract search filters. This presents a potential surface where malicious instructions embedded in the user's description could attempt to influence the agent's behavior or tool parameters.
Ingestion points: The natural language ICP description is ingested via $ARGUMENTS in SKILL.md.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present when processing the input.
Capability inventory: The skill possesses capabilities to search for organizations, perform bulk enrichment of people and companies (including PII), and create contacts in a CRM via the apollo_contacts_create tool.
Sanitization: No explicit sanitization or validation of the extracted search filters is defined before they are passed to the MCP tools.
Sensitive Data Handling (PII): The skill is designed to retrieve and display personal information, including emails and phone numbers, through the apollo_people_bulk_match tool. While this aligns with the skill's intended prospecting purpose, it involves the handling and presentation of sensitive contact data.

prospect