The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses Bash to execute composio CLI tool commands, specifically search for discovery and execute for running verification tasks. This is the standard method for using this vendor's toolset.
[INDIRECT_PROMPT_INJECTION]: The skill accepts untrusted data from user arguments and external platforms which is then interpolated into shell commands.
Ingestion points: User input through the $ARGUMENTS variable and data retrieved from Gorgias or HubSpot contact records in Step 2.
Boundary markers: Absent; input is placed directly into JSON payloads within command-line strings.
Capability inventory: Execution of shell commands via the composio tool, which has the ability to interact with external APIs.
Sanitization: No explicit validation or sanitization of the email strings is mentioned in the instructions before they are passed to the shell environment.
[DATA_EXFILTRATION]: The skill is designed to send email addresses to external verification services (Clearout, Mailcheck). While this is the primary functionality, it involves transmitting data from the agent's context to third-party endpoints.

email-verify