The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill implements currency detection and conversion logic that ingests untrusted data from browser cookies and CDN headers to influence sensitive operations such as payment processing.
Ingestion points: The detectCurrency function in SKILL.md directly reads from req.cookies.preferred_currency and req.headers['cf-ipcountry'] to determine the active currency.
Boundary markers: The skill lacks explicit boundary markers or validation logic to ensure that user-provided currency strings match an allowlist before being used in downstream logic.
Capability inventory: The skill demonstrates capabilities for network operations (fetching rates from OpenExchangeRates, creating Stripe PaymentIntents) and interacting with a Redis cache.
Sanitization: While the code lowercases the currency for Stripe, the targetCurrency used to access the exchange rates object (rates[targetCurrency]) is not sanitized, which could lead to unexpected behavior if malicious property names are provided.

multi-currency