The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to process and interpret external 'order forms' or 'PO documents' to configure billing and entitlements via the Stigg API. This data ingestion creates an attack surface for indirect prompt injection where malicious instructions inside a document could influence the agent's actions.
Ingestion points: Customer order forms and purchase orders are ingested as described in references/billing-and-invoices.md and references/pricing-models.md.
Boundary markers: No specific delimiters or 'ignore embedded instructions' warnings are suggested for processing the content of these external documents.
Capability inventory: The skill utilizes curl to perform POST, PATCH, and DELETE operations on the api.stigg.io endpoint, allowing for significant state changes in the billing system.
Sanitization: No explicit validation or sanitization of the extracted data is required before it is used to populate API requests.

stigg-contracts