quickbooks-online-api
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents the implementation of webhooks and Change Data Capture (CDC), creating a surface where untrusted data from the external QuickBooks API could influence agent behavior if not properly handled.
- Ingestion points: SKILL.md contains code examples for a Node.js webhook listener and a Python CDC sync function that ingest JSON payloads from external endpoints.
- Boundary markers: The provided code snippets do not implement specific delimiters or 'ignore instructions' markers for the data fields received from the API.
- Capability inventory: The skill's scripts utilize network capabilities via
requests,axios, andfetchto perform API operations. - Sanitization: The examples demonstrate processing incoming fields (e.g.,
entity.name,entity.id) without explicit sanitization or validation logic.
Audit Metadata