The Agent Skills Directory

[SAFE]: The skill focuses on architectural best practices for Flutter development, specifically regarding error handling and observability.
[DATA_EXPOSURE]: The instructions explicitly warn against logging sensitive data such as passwords, tokens, and PII. It provides a redact function example in reference/anti-patterns.md to demonstrate how to strip sensitive information before it reaches logs or crash reporters.
[INDIRECT_PROMPT_INJECTION]: The skill defines a clear boundary for processing external data (exceptions, response bodies, and stack traces). It mandates mapping these to internal, sealed Failure objects at the data layer, which effectively prevents untrusted technical details from being directly rendered or processed in the UI layer without sanitization.
[REMOTE_CODE_EXECUTION]: There are no patterns of remote code execution. References to third-party services like Sentry and Firebase Crashlytics are restricted to standard observability and crash reporting contexts.

error-handling