Enterprise Architecture Principles

The LocalM AiD Framework defines 27 principles across 7 categories for operating, managing, and securing AI-assisted software development.

Total Principles 27

Categories 7

Scope AI Tool Operations

Status 🔍 Under Peer Review

Scope: These principles govern how to operate AI coding tools safely — NOT general software engineering practices. See Mandate for scope definition.

Framework Scope

✅ IN SCOPE	❌ OUT OF SCOPE
AI tool configuration	Software design patterns
Agent permissions/sandboxing	Architecture styles
Git access controls for AI	General testing methodology
Data classification for AI	DevOps practices (general)
Audit trails & compliance	Coding standards (general)
Autonomy level governance	Language/framework selection

Applicable Tool Categories:

IDE-integrated AI assistants
Terminal-based AI agents
Agentic coding environments
Model Context Protocol (MCP) servers
Agent-to-Agent (A2A) protocols
Code completion and suggestion tools

Principles Architecture

flowchart TB
    subgraph Tenets["CORE TENETS"]
        T1["1. Human Agency"]
        T2["2. Structured Interaction"]
        T3["3. Continuous Validation"]
        T4["4. Traceability"]
        T5["5. Progressive Maturity"]
    end

    subgraph Categories["7 CATEGORIES"]
        PS["PS"]
        TSI["TSI"]
        TTA["TTA"]
        DC["DC"]
        TQC["TQC"]
        DM["DM"]
        GSC["GSC"]
    end

    subgraph Principles["27 PRINCIPLES"]
        Structure["Each principle follows TOGAF structure:<br/>• Statement - What the principle declares<br/>• Rationale - Why it matters<br/>• Implications - How it affects development<br/>• Maturity - Requirements by level<br/>• Governance - Compliance measures"]
    end

    Tenets --> Categories
    Categories --> Principles

Categories Overview

Code	Category	Principles	Focus
PS	Planning & Strategy	4	AI tool adoption planning
TSI	Tool Selection & Integration	3	Tool evaluation & MCP governance
TTA	Team Training & Adoption	3	AI tool proficiency
DC	Development & Coding	4	AI interaction & autonomy
TQC	Testing & Quality Control	4	AI output validation
DM	Deployment & Maintenance	2	AI pipeline controls
GSC	Governance, Security & Compliance	7	Permissions, sandboxing, git controls

PS: Planning & Strategy

Focus: AI tool adoption planning, capability assessment, maturity roadmap

ID	Principle Name	Statement Summary
PS-001	AI Operations Planning	Plan AI tool deployment before adoption
PS-002	Strategic AI Integration	Align AI adoption with business strategy
PS-003	Risk-Based Planning	Assess AI operational risks upfront
PS-004	Structured Communication	Establish prompt and interaction standards

View PS Principles →

TSI: Tool Selection & Integration

Focus: AI tool evaluation criteria, approval process, MCP/protocol governance

ID	Principle Name	Statement Summary
TSI-001	Evaluation Framework	Standardize AI tool evaluation criteria
TSI-002	Integration Standards	Define integration requirements for AI tools
TSI-003	Protocol Adoption	Govern MCP servers and A2A protocol usage

View TSI Principles →

TTA: Team Training & Adoption

Focus: Building team capability for AI-assisted development

ID	Principle Name	Statement Summary
TTA-001	Skills Development	Structured training for AI tool proficiency
TTA-002	Adoption Governance	Manage organizational AI adoption responsibly
TTA-003	Knowledge Sharing	Establish mechanisms for AI practice sharing

View TTA Principles →

DC: Development & Coding

Focus: AI interaction modes, agent collaboration, autonomy level governance

ID	Principle Name	Statement Summary
DC-001	AI-Human Collaboration	Human oversight of AI code generation
DC-002	Prompt Engineering	Standardize effective prompt practices
DC-003	Code Attribution	Track and attribute AI-generated code
DC-004	Agentic Development	Govern autonomous AI agent operations

View DC Principles →

TQC: Testing & Quality Control

Focus: AI output validation, security scanning of AI-generated code

ID	Principle Name	Statement Summary
TQC-001	AI Output Validation	Validate AI-generated code before use
TQC-002	Security Scanning	Security validation of AI-generated code
TQC-003	Quality Gates	Quality checkpoints for AI outputs
TQC-004	Continuous Validation	Ongoing validation throughout AI lifecycle

View TQC Principles →

DM: Deployment & Maintenance

Focus: AI-specific pipeline controls, AI tool monitoring, incident response

ID	Principle Name	Statement Summary
DM-001	AI Pipeline Gates	AI-specific quality gates in CI/CD
DM-002	AI Operations	Monitor AI tool usage and operations

View DM Principles →

GSC: Governance, Security & Compliance

Focus: Permissions, sandboxing, git controls, data classification, audit trails

ID	Principle Name	Statement Summary
GSC-001	Governance Framework	Establish AI development governance
GSC-002	Permission Boundaries	Define AI agent permission limits
GSC-003	Sandboxing Requirements	Isolate AI tool execution environments
GSC-004	Git Access Controls	Restrict AI agent git operations
GSC-005	Data Classification	Control data exposure to AI tools
GSC-006	Audit & Compliance	Maintain AI activity audit trails
GSC-007	Responsible AI	Ethical AI implementation practices

View GSC Principles →

Principle Structure

Each principle follows the TOGAF-aligned structure with LocalM AiD extensions:

flowchart TB
    Statement["📋 STATEMENT<br/><i>What the principle declares - clear, concise, actionable</i>"]
    Rationale["💡 RATIONALE<br/><i>Why it matters: Business Value, Technical Foundation, Risk</i>"]
    Implications["⚙️ IMPLICATIONS<br/><i>Development, Governance, Skills, Tools requirements</i>"]
    Maturity["📊 MATURITY ALIGNMENT<br/><i>[LocalM AiD Extension]</i><br/>Requirements at Base (L1), Medium (L2), High (L3)"]
    Governance["🔒 GOVERNANCE<br/><i>[LocalM AiD Extension]</i><br/>Compliance checklists, exception processes, audit"]

    Statement --> Rationale --> Implications --> Maturity --> Governance

Maturity Integration

flowchart LR
    subgraph L1["BASE (L1)<br/>Foundation"]
        L1C["Core requirements<br/>Human oversight"]
    end

    subgraph L2["MEDIUM (L2)<br/>Enhanced"]
        L2C["Expanded automation<br/>Structured governance"]
    end

    subgraph L3["HIGH (L3)<br/>Advanced"]
        L3C["Full AI-assisted<br/>Predictive optimization"]
    end

    L1 --> L2 --> L3

Each principle defines specific requirements at each level. Organizations implement based on their current maturity.

Level	AI Autonomy	Focus
Base (L1)	AI-Assisted Options	Minimum requirements; core practices
Medium (L2)	AI-Assisted Selection	Extended automation; structured governance
High (L3)	Partial Automation	Advanced/autonomous capabilities