Skip to content

agents.md

Latest commit

 

History

History
129 lines (102 loc) · 5.2 KB

agents.md

File metadata and controls

129 lines (102 loc) · 5.2 KB
description The Meta-Intelligence & Vibe Coding Protocol defining how AI agents interact with the repository and follow architectural guidelines.
tags
vibe coding
ai agents
execution protocol
context mapping
global constraints

AGENTS.md: The Meta-Intelligence & Vibe Coding Protocol

🌍 1. Global Vision

The Gold Standard for AI-Driven Development Instructions.

This repository is not a codebase; it is a Machine-Readable Intelligence Layer. Its mission is to bridge the gap between human intent and AI execution by providing a curated ecosystem of meta-instructions. We aim to:

  • Automate the Mundane: Shift the burden of boilerplate and syntax to the AI.
  • Architectural Excellence: Force AI agents to respect high-level patterns (FSD, MVC) before writing a single line of logic.
  • Vibe Coding Synchronization: Provide a unified "Source of Truth" for any LLM-based agent (Cursor, Windsurf, Copilot) to ensure "perfect first-time" generation.

🗺️ 2. Context Mapping

AI Agents MUST traverse this repository using the following hierarchical logic:

graph TD
    Global[Global Constraints: .agents/rules/github-seo.md] --> Arch[Architectural Patterns: /architectures/]
    Arch --> FSD[Feature-Sliced Design]
    Arch --> MVC[Model-View-Controller]
    
    Global --> Tech[Tech Stack Specifics]
    Tech --> FE[Frontend: /frontend/]
    Tech --> BE[Backend: /backend/]
    
    FE --> TS[TypeScript/JS]
    FE --> Angular[Angular]
    
    BE --> Nest[NestJS]
    BE --> Express[Express]

    style Global fill:#f9f,stroke:#333,stroke-width:4px
    %% Added Design Token Styles for Mermaid Diagrams
    classDef default fill:#e1f5fe,stroke:#03a9f4,stroke-width:2px,color:#000;
    classDef component fill:#e8f5e9,stroke:#4caf50,stroke-width:2px,color:#000;
    classDef layout fill:#f3e5f5,stroke:#9c27b0,stroke-width:2px,color:#000;

    class BE component;
    class Tech component;
    class MVC component;
    class TS component;
    class Angular component;
    class FSD component;
    class FE component;
    class Express component;
    class Arch component;
    class Nest component;
Loading

Resource Directory

Path Intent Agent Constraint
/.agents/rules/ Global SEO & Metadata Must be injected into the system prompt for project indexing.
/architectures/ System Design Blueprints Defines the folder structure and module boundaries.
/frontend/ UI/UX & Client Logic Specific rules for React, Angular, and TypeScript safety.
/backend/ Server-side & API Rules for scalability, security, and DTO validation.

🛠️ 3. Tool-Specific Integration

To leverage these rules in your IDE, follow these implementation paths:

🧩 Cursor AI (.cursorrules)

Copy the content of the relevant .md file (e.g., architectures/feature-sliced-design/readme.md) into your project's root .cursorrules. Pro-tip: Use @AGENTS.md in Cursor Chat to initialize the session context.

🏄 Windsurf (.windsurfrules)

Reference this repository in your .windsurfrules file:

{
  "ruleDirs": ["path/to/best-practise/architectures/fsd"],
  "active": true
}

💻 VS Code (Custom Instructions)

For extensions like Cloud Code or GitHub Copilot, append the contents of AGENTS.md to your Settings > Custom Instructions field to establish the "Architectural Guardrails."

🤖 4. Agent Execution Protocol

Any AI Agent interacting with this repository MUST follow this 4-step execution loop:

  1. Context Discovery: Scan AGENTS.md to identify the project's meta-structure.
  2. Constraint Identification: Locate the specific .md file corresponding to the user's task (e.g., if the user asks for a NestJS service, read backend/nestjs/readme.md).
  3. Hierarchy Alignment: Prioritize architectures/ constraints over frontend/ or backend/ specifics to ensure structural integrity.
  4. Verification: Before outputting code, run a "Constraint Check" against the identified rules to ensure zero violations of the "Best Practise" standard.

🔍 5. SEO & Discoverability

Tag Purpose
#VibeCoding Optimized for high-velocity, intent-based programming.
#CleanCode Enforces industry-standard patterns automatically.
#AgenticWorkflow Designed for autonomous and semi-autonomous AI agents.

Language - English

All documentation and code comments must be in English.

  1. Technical Easy English language
  2. No slang or colloquialisms
  3. No jargon or technical terms without explanation
  4. No abbreviations or acronyms without explanation
  5. No metaphors or analogies without explanation
  6. No sarcasm or irony
  7. No humor or jokes
  8. No cultural references
  9. No idioms or expressions
  10. No slang or colloquialisms

Status: READY_FOR_INFERENCE Version: 1.0.0 Architect: Senior AI Solutions Architect