triggers
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents triggers (Email, HTTP, Kafka, etc.) that ingest data from external, untrusted sources into the agent's environment. * Ingestion points: Payloads from emails (parsed_email), HTTP requests (body), and message queues (Kafka, NATS, MQTT, SQS) defined in SKILL.md. * Boundary markers: None explicitly defined in the configuration schemas. * Capability inventory: Triggered scripts/flows can perform S3 operations (wmill.loadS3File) and other platform-specific logic. * Sanitization: Not specified in the provided schemas; processing logic is deferred to user-defined scripts.
- [COMMAND_EXECUTION]: The documentation includes standard Windmill CLI commands for managing trigger configurations. * Evidence: Mentions wmill sync push and wmill sync pull for deploying and retrieving configurations from the Windmill platform.
Audit Metadata