The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the configuration of an external MCP server endpoint at https://rube.app/mcp. This is a non-standard service that is not part of the trusted vendors list.
[INDIRECT_PROMPT_INJECTION]: The skill has a significant attack surface for indirect prompt injection through external data sources.
Ingestion points: The skill ingests untrusted data via SENDGRID_IMPORT_CONTACTS (CSV imports), SENDGRID_CREATE_SINGLE_SEND (HTML and plain text email content), and SENDGRID_ADD_OR_UPDATE_A_CONTACT (contact names and metadata).
Boundary markers: There are no instructions to use delimiters or ignore embedded commands within the processed data.
Capability inventory: The skill possesses high-impact capabilities including creating marketing campaigns (SENDGRID_CREATE_SINGLE_SEND), managing contacts, and deleting lists (SENDGRID_REMOVE_LIST_AND_OPTIONAL_CONTACTS).
Sanitization: There are no instructions for escaping or validating external content before it is processed by the agent.
[METADATA_POISONING]: The skill's frontmatter contains a risk: critical field. While potentially a genuine warning, such labels in metadata can be used to influence the agent's behavior or safety filtering in a non-standard way.

sendgrid-automation