The Agent Skills Directory

[SAFE]: The skill outlines safe coding practices and implementation patterns for Nango actions and syncs, focusing on schema validation and platform constraints.
[SAFE]: Instructions explicitly forbid the inclusion of arbitrary third-party packages, limiting imports to approved built-in modules like zod, crypto, and url to minimize supply chain risks.
[SAFE]: The skill identifies and warns against operations that are incompatible with Nango's sandboxed runtime, such as direct file system (fs) access or large binary transfers within actions, directing users toward more appropriate proxy script patterns instead.
[SAFE]: Security best practices are encouraged through the mandatory use of Zod for input/output validation and the recommendation to explicitly manage HTTP retries to prevent unintended recursion or service exhaustion.

building-nango-functions