The Agent Skills Directory

[SAFE]: The skill provides instructions for using the 'eve' CLI tool to diagnose deployment and job failures. All commands (e.g., eve system health, eve job diagnose, eve env diagnose) are consistent with the skill's stated purpose of troubleshooting.
[SAFE]: Instructions for managing secrets and authentication follow standard development practices. The skill correctly advises using the tool's built-in commands (eve secrets list, eve auth login) rather than manual or insecure methods.
[SAFE]: The skill demonstrates a defensive approach to external data. When processing Slack alerts, it explicitly instructs the agent to verify the information using the eve env diagnose command rather than acting on the alert text alone.
[SAFE]: No obfuscation, prompt injection, or suspicious network operations were found. All mentioned resources, such as labels for GitHub and registry host configurations, are standard metadata for containerized deployments.

eve-troubleshooting