The Agent Skills Directory

[SAFE]: No malicious code, unauthorized network activity, or credential harvesting patterns were detected. The skill uses official RevenueCat infrastructure for its operations.
[PROMPT_INJECTION]: The directives regarding the agent's communication (e.g., not stating that RevenueCat cannot filter customers) are legitimate persona and capability guidelines for a tool-specific skill and do not attempt to bypass safety filters or ignore system instructions.
[INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing potentially untrusted customer data retrieved from RevenueCat APIs.
Ingestion points: Customer profile data such as app user IDs and emails ingested via the preview-audience tool.
Boundary markers: The instructions do not specify explicit delimiters for data presentation, relying on standard agent output formats.
Capability inventory: The skill's tools are limited to RevenueCat API actions; no arbitrary command execution or external network capabilities are available.
Sanitization: The skill does not instruct the agent to sanitize customer-supplied fields, though the risk is considered minimal within the context of the RevenueCat dashboard functionality.

revenuecat-audiences