Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/xxyoudeadpunkxx/sensecraftxstudio/llms.txt

Use this file to discover all available pages before exploring further.

SensecraftXStudio gives AI coding agents a clear operating contract before they inspect files, change code, or report that work is complete. It is a single reusable file — AGENTS.md — that you copy into any project to enforce disciplined, bounded, and reviewable AI behavior throughout a session.

Introduction

Understand what SensecraftXStudio is, what problem it solves, and who it is for.

Quickstart

Copy AGENTS.md into your project and activate the contract in under a minute.

The Contract

Explore every section of the AGENTS.md operating contract in detail.

Guides

Learn how to adopt, run, and get the most out of SensecraftXStudio sessions.

​
What it does

When an AI assistant reads AGENTS.md and treats it as its operative frame, it is required to:
  • Close context before acting — identify the real target, source of authority, and operating mode before any move.
  • Surface scope expansion — name any perimeter crossing before proceeding, not after.
  • Execute minimally — choose the smallest correct intervention, avoiding opportunistic cleanup or restructuring.
  • Report honestly — finish consequential work with a structured account of what changed, what is verified, what is inferred, and whether the task is complete.
1

Copy AGENTS.md into your project root

Download or copy the canonical AGENTS.md file to the root of the repository the assistant will work in.
2

Start a fresh assistant session

Open a new session with your AI assistant and grant it access to the workspace where AGENTS.md lives.
3

Activate the contract

Give the assistant this single instruction before any task:
Read AGENTS.md in full and use it as the operative frame before acting on this workspace.
4

Give the real task

Assign the actual work. The contract remains active for the entire session and governs how every move is framed, executed, and reported.
AGENTS.md is not background documentation to read once and forget. It is an active contract that must remain the operative frame for the full session. See Contract Activation for details on drift and re-entry.

​
Key concepts

Horizontal Plane

Three axes — Close, Read, Execute — that structure every move.

Stop Conditions

When the assistant must halt and surface a conflict before proceeding.

Final Response Contract

Touch, Ground, State, and Convergence fields for honest task closure.

Build docs developers (and LLMs) love

Get started for freeTalk to us