The Agent Skills Directory

[SAFE]: The skill provides documentation and code examples for Phoenix AI observability. All external references and dependencies are standard libraries managed by Arize AI for OpenInference tracing.
[DATA_EXFILTRATION]: While the skill involves sending trace data to a Phoenix server, this is the explicitly stated primary purpose. The documentation includes dedicated guides for production safety, such as production-python.md and production-typescript.md, which provide instructions for PII masking (e.g., using OPENINFERENCE_HIDE_INPUTS environment variables) to prevent sensitive data exposure.
[EXTERNAL_DOWNLOADS]: The skill references standard package managers like pip and npm for installing official Phoenix and OpenInference libraries. No suspicious or unverified download sources or piped remote execution patterns were found.
[CREDENTIALS_UNSAFE]: Security best practices for secret management are followed. Documentation examples consistently use environment variables for API keys rather than hardcoding sensitive credentials.

phoenix-tracing