ido4 — How It Works

2 senior humans + AI agents + ido4 governance = the output of a 10-person team.
ido4 gives AI agents shared understanding, long-term coherence, and deterministic quality enforcement.

Humans
PM + Tech Architect
Product vision, system design, strategic decisions, oversight
ido4shape
The WHAT
Conversation with stakeholders → strategic spec. What to build and why.
ido4 MCP
Context + Governance Engine
Gives every AI session full project understanding. Validates every action. Tracks everything.
Context Assembly Task Intelligence BRE Validation Audit Trail Compliance Scoring Work Distribution
GitHub
Living Specifications
Issues are not tickets — they're context carriers encoding intent, criteria, and position.
AI Agents
Execution
Claude Code sessions. Read specs, build, test, commit. Governed by ido4.
strategic spec →
decomposes into tasks →
reads & writes issues
context + next task
validates transitions
Continuous Loop
Agent reads context → picks task → develops → transitions through BRE → writes structured context → next agent reads accumulated knowledge. Every session inherits full project understanding.
Development Governance Platform — deterministic methodology enforcement for AI-augmented teams