async-messaging
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFE
Full Analysis
- [NO_CODE]: The skill consists entirely of instructional Markdown documentation and a YAML configuration file. It does not include any executable scripts, binaries, or tool definitions.
- [SAFE]: The skill provides strict architectural constraints and 'Stop Conditions' designed to prevent the AI from giving unsafe production advice (e.g., preventing infinite retry loops or non-idempotent side effects).
- [EXTERNAL_DOWNLOADS]: References several external sources including official documentation for AWS, Apache Kafka, RabbitMQ, Stripe, and standards organizations (IETF, ACM). All referenced domains are well-known, trusted technology services or academic repositories.
- [PROMPT_INJECTION]: The skill uses strong imperative language ('Domain Law', 'MUST', 'NEVER') to define its operating boundaries, but these are functional constraints intended to improve engineering quality rather than attempts to override the underlying agent's safety filters.
- [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or secrets were found. The documentation correctly advises developers to manage secrets via configuration and environment variables.
Audit Metadata