Humsana FAQs

Question 1

Can I customize Humsana's safety thresholds and behavior?

Accepted Answer

Yes, Humsana is highly configurable. You can adjust fatigue thresholds, AI rewrite warning/block limits, add custom dangerous command patterns, and switch between `dry_run` and `live` execution modes via a local YAML configuration file.

Question 2

Is Humsana privacy-friendly?

Accepted Answer

Absolutely. Humsana is 100% local-first. All your data stays on your machine, with no network calls (except for an optional webhook), no telemetry, and fully auditable open-source code.

Question 3

What types of actions does Humsana protect against?

Accepted Answer

It offers 'Command Protection' to prevent dangerous shell commands (e.g., `rm -rf`) and 'AI Rewrite Protection' to safeguard against AI-generated changes that delete or modify a significant number of lines, especially when fatigued.

Question 4

How does Humsana detect developer fatigue?

Accepted Answer

Humsana's local daemon gathers behavioral data to assess your cognitive state, including fatigue levels. When your fatigue reaches a configurable threshold, it intervenes to protect against critical errors.

Question 5

What is Humsana and how does it help developers?

Accepted Answer

Humsana provides cognitive security for AI-assisted development. It acts as a safety interlock, preventing you from executing dangerous commands or accepting large, potentially harmful AI code rewrites when it detects you're fatigued.

Humsana

Humsana

주요 기능

사용 사례

주요 기능

사용 사례