The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill defines patterns for ingesting external data (invoices, bills, journal entries) from Visma Netvisor, creating a potential surface for indirect injection if that data contains malicious instructions.
Ingestion points: Data retrieved via apideck.accounting.invoices.list and related methods as described in SKILL.md.
Boundary markers: Not explicitly defined in the instructional text; the agent relies on its own parsing logic.
Capability inventory: The skill documents the use of the Apideck SDK to perform network requests to unify.apideck.com.
Sanitization: No specific sanitization logic is provided in the instructions; it is assumed the unified API handles data normalization.
[SAFE]: Credential management in examples correctly utilizes process.env for sensitive keys, preventing hardcoded secrets.
[SAFE]: All external URL references target official domains belonging to the author (Apideck) or the integrated service (Visma Netvisor).

visma-netvisor