exact-online-nl
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
- [DATA_EXFILTRATION]: Performs authenticated network requests to Apideck's official API endpoints at unify.apideck.com. These operations are essential for the skill's primary function and target the vendor's established infrastructure using secure credential handling.
- [EXTERNAL_DOWNLOADS]: References official OpenAPI specifications and developer documentation from trusted vendor domains, including specs.apideck.com and developers.apideck.com.
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external Exact Online NL records, which creates a potential surface for indirect prompt injection.
- Ingestion points: Accounting API responses for resources such as invoices, bills, and payments (SKILL.md).
- Boundary markers: No specific delimiters or boundary warnings are defined in the instructions for separating external data from agent commands.
- Capability inventory: The agent can list, create, and update accounting transactions, and can also perform arbitrary requests to the Exact Online NL API via a proxy endpoint (SKILL.md).
- Sanitization: No explicit sanitization or validation logic is described for the retrieved accounting data before it is incorporated into the agent context.
Audit Metadata