Skip to content

AGENTS.md

Latest commit

 

History

History
33 lines (21 loc) · 1.47 KB

AGENTS.md

File metadata and controls

33 lines (21 loc) · 1.47 KB

🚨 CRITICAL SYSTEM INSTRUCTION 🚨

Before answering ANY prompt or executing ANY code in this repository, you MUST first read and strictly adhere to ALL the rules defined in the following directory:

Rules Location

📁 .agent/rules/

🚨 CRITICAL TOKEN LIMIT PROTOCOL 🚨

Do NOT load all files in .agent/rules/ simultaneously. You will exceed the context window.

You MUST follow this strict loading sequence:

  1. Load Core Fundamentals: Always read 01-general.md through 12-docs.md.
  2. Consult the Lazy-Loading Router: Read .agent/rules/00-index.md to discover technology-specific rules.
  3. Load On Demand: Based on the index, load ONLY the specific .md files relevant to your current task (e.g., react.md for a React task).

Why This File Exists

This project uses a unified rule system to ensure consistent AI behavior across all AI-powered IDEs and tools (Cursor, Windsurf, GitHub Copilot, Cline, Claude, Gemini, Trae, Roo Code, Augment, Amazon Q, Kiro, Continue, Junie, etc.). This file is a redirect entry point — the actual rules live in .agent/rules/ as the Single Source of Truth.

Failure to follow the rules inside .agent/rules/ and the Lazy-Loading protocol is completely unacceptable.

Slash Commands / Workflows

All AI slash commands and standard operating procedures (SOPs) are located in .agent/workflows/. If the user invokes a command like /speckit.plan, execute the corresponding markdown file in that directory.