The Agent Skills Directory

[COMMAND_EXECUTION]: The skill employs curl and jq to interact with API endpoints and process data. It also references a platform-specific troubleshooting tool, zero, to verify environment configurations.
[EXTERNAL_DOWNLOADS]: Network operations are conducted with api.sendgrid.com. As this is a well-known service providing the core functionality of the skill, it is documented as a legitimate and expected destination.
[PROMPT_INJECTION]: The skill processes untrusted input such as recipient information and email bodies, which represents a surface for indirect prompt injection.
Ingestion points: User-supplied email content and marketing contact data.
Boundary markers: None implemented in the provided workflows.
Capability inventory: Writing to the local filesystem (/tmp/sendgrid_request.json) and making network requests via curl.
Sanitization: No explicit validation or filtering of the user-provided data is defined.

sendgrid