The Agent Skills Directory

[SAFE]: The skill uses a local websocket connection (localhost) for its engine connection example, which is a safe practice for development environments.
[SAFE]: No hardcoded credentials, remote code downloads, or obfuscation techniques were detected in the skill content.
[SAFE]: The skill documents standard architectural patterns for building event-driven workers using the iii engine SDK.
[PROMPT_INJECTION]: Analysis of indirect prompt injection surface:
Ingestion points: SKILL.md defines triggers (HTTP, queue, stream, subscribe) that process data from external event sources.
Boundary markers: None are explicitly defined in the provided implementation patterns.
Capability inventory: Function handlers defined via registerFunction and cross-worker invocations via trigger() enable execution of logic across different workers and languages.
Sanitization: No specific data validation or sanitization logic is included in the generic patterns, as these are meant to be implemented by the developer based on specific use cases.

functions-and-triggers