The Agent Skills Directory

[SAFE]: The skill provides instructions for AI application observability using the DeepEval SDK. No malicious patterns such as prompt injection, obfuscation, or unauthorized data exfiltration were detected.
[CREDENTIALS_UNSAFE]: While the skill mentions the use of a CONFIDENT_API_KEY, it correctly instructs users to use environment variables or CLI login rather than hardcoding secrets. It also includes explicit warnings against tracing sensitive information such as API keys and credentials.
[EXTERNAL_DOWNLOADS]: References to the external package deepeval via pip and the Confident AI service are consistent with the vendor's identity and the skill's stated purpose.

deepeval-tracing