finance-billing-ops
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill defines a legitimate operational workflow for revenue analysis and billing verification. There are no malicious commands, obfuscated segments, or attempts to bypass safety protocols.
- [DATA_EXFILTRATION]: While the workflow involves accessing sensitive revenue data and source code repositories, it does so within the intended functional scope of the skill. No evidence of unauthorized data transfer or exfiltration attempts was found.
- [PROMPT_INJECTION]: The skill manages an attack surface for indirect prompt injection by ingesting external data. Ingestion points: Reads customer billing data from Stripe (sales, subscriptions, refunds) and source code from GitHub repositories. Boundary markers: No explicit instructions for delimiters or isolation of untrusted content are provided. Capability inventory: The skill utilizes file reading via github-ops and remediation actions through customer-billing-ops. Sanitization: Input validation or sanitization steps for external data are not explicitly defined in the workflow.
Audit Metadata