The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local diagnostic CLI tool via the $AIONUI_HELPER_BIN environment variable. The commands are restricted to read-only diagnostic operations (e.g., diagnose overview, diagnose health, diagnose logs tail).
[DATA_EXPOSURE]: While the skill accesses logs and configuration summaries, it explicitly instructs the agent to redact and summarize sensitive findings and notes that the CLI itself redacts known secret fields by default.
[SAFE]: The skill follows least-privilege principles by separating troubleshooting from configuration changes (relegated to a separate skill) and using controlled HTTP 'escape hatches' restricted to local API health and metadata endpoints.

aionui-troubleshooting