AI Usage Considerations
The AI assistant significantly improves work efficiency, but as a professional developer tool, we must emphasize safety and risk management alongside convenience.
⚠️ Do Not Let AI Blindly Execute Dangerous Commands
Commands provided by AI are not always 100% correct or absolutely safe. Even if the AI provides repair suggestions, we strictly prohibit the AI from bypassing the user to execute code directly. All code or commands generated by AI, when pushed to the terminal, are only inserted into the command line input box.
You must:
- Carefully read the specific parameters of the command.
- Confirm that the command will not cause damage to the production environment.
- Manually press the Enter key to execute it only when fully understood.
🔒 Privacy and Data Security
We understand the sensitivity of server environments, and Nexa follows a strict data privacy policy when handling AI requests:
- On-Demand Trigger: Nexa's AI will never silently collect or upload any log streams output by your terminal in the background. Content is only processed when you actively highlight it and interact.
- Desensitization Recommendations: When analyzing logs containing user passwords, payment keys, or confidential tokens, we strongly recommend manually desensitizing key information before sending.