The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external data through webhook event handlers (e.g., webEventConsumer.event("chat_message.sent", ...)). This constitutes an ingestion point for untrusted data from the Zoom platform.
Ingestion points: Webhook payloads processed in examples/getting-started-pattern.md and examples/multi-client-pattern.md.
Boundary markers: No specific delimiters or "ignore instructions" warnings are present in the provided code snippets.
Capability inventory: The skill allows for REST API operations through client.endpoints.* wrappers.
Sanitization: The provided examples do not explicitly demonstrate sanitization of event payloads, which is a standard risk for skills processing external messaging data.
[COMMAND_EXECUTION]: The skill provides patterns for starting a Node.js server instance using client.start(), which binds the application to a local network port to receive webhooks.
[EXTERNAL_DOWNLOADS]: The skill references the official Zoom Rivet SDK package (@zoom/rivet) and provides links to official documentation and sample repositories on GitHub. All external references target trusted vendor domains or well-known services associated with the skill author.

rivet-sdk