freshbooks
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from FreshBooks (such as invoices, bills, and customers). This represents a theoretical surface for indirect prompt injection if data retrieved from the FreshBooks API contains malicious instructions intended to influence the agent's behavior.
- Ingestion points: Data is ingested through calls to the FreshBooks API via Apideck (e.g.,
apideck.accounting.invoices.list). - Boundary markers: None explicitly defined in the instruction templates; reliance is placed on the agent's underlying safety filters.
- Capability inventory: The skill allows network operations to Apideck's Unified API and direct FreshBooks API access via the Proxy endpoint.
- Sanitization: Standard SDK usage is demonstrated without explicit manual sanitization of retrieved strings.
- [EXTERNAL_DOWNLOADS]: The skill references the official
@apideck/unifypackage for Node.js. It also includes links to documentation and OpenAPI specifications hosted onapideck.com,freshbooks.com, andgithub.com. These are recognized as legitimate resources belonging to the vendor and the integrated service.
Audit Metadata