v0.0.1 • In Peer Review

Reference Documentation

Technical reference materials for the LocalM™ AiD framework.

TYPE Technical Reference
AUDIENCE Implementers & Architects

Principle Quick Reference

By Category

Category Code Principles Focus
Planning & Strategy PS 4 Strategic AI adoption
Tool Selection & Integration TSI 3 Tool evaluation & integration
Team Training & Adoption TTA 3 Skills development
Development & Coding DC 6 Day-to-day AI usage
Testing & Quality Control TQC 3 Quality assurance
Deployment & Maintenance DM 2 Operational concerns
Governance, Security & Compliance GSC 10 Security & governance

By Maturity Level

Level Name Description Suitable For
L1 Foundation Basic governance, manual oversight New adopters
L2 Enhanced Intermediate automation, team-wide practices Growing teams
L3 Advanced Advanced automation, minimal supervision Mature organizations

Standards Reference

AGENTS.md Standard

Project-level instructions for AI agents defining:

  • Allowed operations
  • Forbidden actions
  • File access patterns
  • Communication protocols

Related Principles: DC-005, DC-006, GSC-002

SKILL.md Standard

Modular capability definitions for AI agents:

  • Skill declaration
  • Input/output contracts
  • Dependencies
  • Usage guidelines

Related Principles: DC-005, TSI-002

Model Context Protocol (MCP)

Anthropic’s standard for AI-tool integration:

  • Tool registration
  • Context passing
  • Response handling
  • Security boundaries

Related Principles: TSI-002, TSI-003, GSC-003

Agent-to-Agent (A2A)

Google’s protocol for inter-agent communication:

  • Agent discovery
  • Task delegation
  • Result aggregation
  • Error handling

Related Principles: TSI-003, DC-006


TOGAF Alignment

LocalM™ AiD principles follow TOGAF’s architecture principle structure:

TOGAF Component LocalM™ AiD Implementation
Statement Clear, actionable principle statement
Rationale 4-dimension rationale table
Implications What changes when applied
+ Maturity Alignment (L1/L2/L3 requirements)
+ Governance (Compliance measures)
+ Related Principles (Cross-references)

MCP (Model Context Protocol)

Standard for agent-to-tool communication. See Protocol Integration principle.

A2A (Agent-to-Agent Protocol)

Standard for multi-agent coordination. See Protocol Integration principle.