# Sector8

> The model proposes. The gate authorises.

AI agents cannot approve themselves. Sector8 places an independent decision gate between agents and real-world actions, so every tool call, API call, model call, and handoff can be allowed, denied, or audited before execution.

REST evaluate returns ALLOW or DENY. If the action cannot be allowed safely, it is denied.

This file is plain text. You do not need JavaScript. Prefer these markdown URLs over the HTML pages.

Contact: hello@sector8.ai
Shadow-mode pilot: https://sector8.ai/contact

Customer docs (product contract): https://sector8.mintlify.app/llms.txt

## Pages

- [Home](/llms/home.md): The model proposes. The gate authorises. Shadow-mode pilot is the front door.
- [Product](/llms/product.md): The pre-execution gate: propose, decide, prove.
- [Shadow Mode](/llms/shadow-mode.md): Two weeks, enforcement off, a report of what the agents already tried.
- [Reason Codes](/llms/reason-codes.md): Public denial codes the gate emits.
- [Security](/llms/security.md): Pre-execution control. Tenant-isolated policy, EU-hosted, encryption, evidence.
- [About](/llms/about.md): An agent cannot approve itself. Founders and mission.
- [Zero Trust for AI Agents](/llms/zero-trust-ai-agents.md): Same Gate product. Verify every governed action before it executes.
- [Contact](/llms/contact.md): Request a shadow-mode pilot. hello@sector8.ai

## Machine-readable

- [reason-codes.json](/reason-codes.json): the ten public codes, with meaning, example, and evidence fields.
- [security.txt](/.well-known/security.txt)

## Full text

- [llms-full.txt](/llms-full.txt): Home, Product, Shadow Mode, Reason Codes, Security, About, Zero Trust, and Contact in one file.

## Also

- [Docs](https://sector8.mintlify.app/guides/introduction)