Codokey Reference Brief
AI Code Access Governance
Reference model document — not a formal certification, audit, or product specification. This is the root definition page for the discipline at codokey.com/category/
Canonical Definition
AI Code Access Governance is the discipline of governing what AI agents, AI-touched code, and agentic development workflows can access, expose, change, and move toward production.
AI coding did not merely change who writes software.
It changed where authority begins.
It changed where authority begins.
Foundational Shift
Before AI agents: Who wrote the code?
After AI agents: What can the code and the agent access, expose, change, and release?
Codokey is the canonical reference model for this category.
Seven Category Layers
01 — AI Agent Capabilities
What can the agent do? Each capability carries distinct governance weight.
02 — Codebase Access
Repository trust boundaries defined before agent execution.
03 — Secrets Boundary
Credential classes excluded from agent read context.
04 — Agent Authority
Total scope of agent permission — minimum privilege default.
05 — Deployment Authority
Production paths require human gates.
06 — Audit Memory
Reconstructable agent action history.
07 — Organizational Accountability
Who owns the governance model and reviews governance debt.
The Codokey Reference Stack
- Charter — why the category exists
- Protocol — how the category is governed
- Scorecard — how posture is assessed
- Cost of Undefined Agent Authority — what failure costs
- Buyer Logic — who structurally needs the category