automation-audit-ops
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and provides a logical framework for auditing existing automation infrastructure without introducing malicious code or unsafe commands.
- [PROMPT_INJECTION]: The instructions do not contain any patterns intended to bypass safety guardrails or override system prompts. The 'Guardrails' section specifically instructs the agent to remain read-only unless prompted otherwise.
- [DATA_EXFILTRATION]: No network requests to external domains or attempts to exfiltrate sensitive workspace data were identified. The audit targets (hooks, workflows, and configs) are handled within the local agent context.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by reading repository scripts, hooks, and configuration files. While this creates an attack surface for indirect prompt injection, it is the primary purpose of the auditing tool and the skill lacks any unsafe interpolation or automatic execution of the ingested content.
Audit Metadata