The Agent Skills Directory

[Data Processing Surface]: The skill is designed to ingest and act upon data from external tools such as get_pricing_context. While this is standard functionality for pricing operations, processing external data always presents a potential surface for indirect prompt injection if that data were to contain malicious instructions.
Ingestion points: The skill reads pricing, cost, margin, and demand signals via the get_pricing_context tool (defined in SKILL.md).
Boundary markers: The instructions do not specify explicit delimiters or "ignore embedded instructions" markers for the data returned by tools.
Capability inventory: The skill utilizes state-changing tools including stage_price_update and stage_promotion to modify store configurations (defined in SKILL.md).
Sanitization: There are no explicit instructions for sanitizing or filtering the content of external pricing data before it is processed by the agent.
Contextual Mitigation: The risk is mitigated by the skill's design, which requires an operator to review and approve a preview before any staged changes are finalized, ensuring human oversight for all financial adjustments.

pricing-promotions