The Agent Skills Directory

[SAFE]: The skill is primarily instructional, providing best practices and templates for narrative systems without executing dangerous commands or making network requests.
[SAFE]: The provided Python snippet for a custom dialogue runner (references/runner.md) correctly implements a secure, whitelisted expression evaluator using the ast module to mitigate potential injection attacks.
[SAFE]: No obfuscation, hardcoded credentials, or unauthorized file access patterns were detected in the skill instructions or reference materials.

dialogue-systems