> ## Documentation Index
> Fetch the complete documentation index at: https://docs.snapsec.co/llms.txt
> Use this file to discover all available pages before exploring further.

# Safety Controls & Guardrails

> Rigorous safety mechanisms preventing downtime, data corruption, or denial-of-service.

Enterprise safety is built into the core of Snapsec AI Pentesting.

***

## Guardrail Safeguards

* **Strict Non-Destructive Operation**: Banned actions include `DROP TABLE`, filesystem deletion, mass emailing, or payload flooding.
* **Automatic Target Health Monitoring**: Continuous heartbeat requests monitor target latency. If application response time exceeds thresholds, the agent automatically throttles back.
* **Human Authorization Gates**: High-risk exploit chains (e.g., administrative password reset) pause for human security analyst approval.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.