iii-low-code-automation
Pass
Audited by Gen Agent Trust Hub on Mar 31, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill defines a pattern for ingesting untrusted external data via webhooks and form submissions, which represents a potential surface for indirect prompt injection.
- Ingestion points:
registerTrigger({ type: 'http' })defined inSKILL.mdfor receiving external webhooks. - Boundary markers: Absent; the provided patterns do not include specific delimiters or instructions to isolate external payloads from the agent's internal logic.
- Capability inventory: The skill utilizes
trigger()to enqueue actions,state::setto persist data, andTriggerAction.Void()for notifications, which are handled by the iii engine worker. - Sanitization: Absent; the reference implementation and patterns do not explicitly provide sanitization or validation steps for the incoming payload data.
Audit Metadata