ConceptDiscovery layer

Rails Baseline Agent Contract Model

Rails Baseline ships AGENTS.md and ARCHITECTURE.md as machine-readable contracts, so the framework treats documentation as the control surface for agent-driven delivery.

By InnovaAI ResearchPublished

What is Rails Baseline Agent Contract Model?

AGENTS.md contracts → predictable agent output

Agent reads AGENTS.md, matches contract, ships feature, agency bills retainer

Rails Baseline ships AGENTS.md and ARCHITECTURE.md as machine-readable contracts, so the framework treats documentation as the control surface for agent-driven delivery. An agency running a $2,500 Rails Baseline Starter Launch (20h setup) can point coding agents at those files before any feature work begins, which keeps tenancy, entitlements, and Pundit authorization patterns consistent across every sprint. The $129 one-time licence covers the source code, 1.x updates, and extension recipes, so the agency pays once and reuses the baseline across multiple client retainers. The model is simple: every new model, limit, or background job gets checked against the documented contracts first, and only then does an agent write code. Agencies that skip this step absorb rework when agents invent their own patterns; agencies that enforce it turn a 20-hour setup into a repeatable delivery template they can resell without rebuilding architecture each time.

ai-code-tools