The Agent Skills Directory

[SAFE]: The skill documents standard orchestration patterns including worker registration, function chaining via queues, and state tracking. It uses the iii engine's built-in primitives for retries and concurrency management.
[INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing external data through workflow payloads.
Ingestion points: Data entering through registerFunction step payloads and HTTP triggers (SKILL.md).
Boundary markers: None identified in the provided patterns to delimit untrusted input from instructions.
Capability inventory: The engine supports enqueuing further steps, updating shared state, and publishing events via the trigger command (SKILL.md).
Sanitization: No explicit sanitization or validation logic is shown for the data payloads processed by the functions.

workflow-orchestration