Core Principle

Scenario-Driven.
Test-First.

One scenario ID traces from the first requirement to the last line of verified code. Tests define "done" before a single line is written. Quality becomes a number — not a feeling.

The Scenario Journey

One ID. Three phases. Complete traceability.

Follow scenario S01 from a one-sentence pain point to automated acceptance.

Phase 1 · WHY S01
"Users need to register with email"
Pain Point P01: No way to create an account
Acceptance GIVEN valid email WHEN submit THEN account created
Same S01
Phase 2 · WHAT S01
Pages Registration form → Success page
Fields email (required, format), password (min 8)
States loading, success, error (duplicate email)
Same S01
Phase 3 · HOW S01
Step 1 Sequence diagram → POST /auth/register
Step 2 OpenAPI YAML + users table DDL
Step 3 12 test cases: UT-S01-001 … ST-S01-012
Step 4 AI generates code + test code
Step 5 openlogos verify → Gate 3.5 PASS ✓
No traceability matrix needed. The scenario ID is the trace. Change a requirement for S01? You know exactly which API, tests, and code are affected.
Test-First Methodology

Define "done" before writing code

This is not traditional TDD. It's designing the entire test system before any implementation.

Traditional TDD
Write one test
↓
Write code to pass it
↓
Refactor
↓
Repeat (code-level thinking)
vs
OpenLogos Test-First
Design all test cases first
↓
Unit tests (function correctness)
Scenario tests (cross-module flow)
API orchestration tests (end-to-end)
↓
Then AI generates code with full context

Why does this matter for AI?

When you tell AI "write a registration feature," it guesses what "done" means. When you tell AI "write code that passes these 12 test cases with IDs UT-S01-001 through ST-S01-012," it has a precise, verifiable target. The output quality difference is dramatic.

Automated Verification

Quality is a number, not a feeling

1

Tests run

npm test / pytest — any framework, any language

2

Results captured

OpenLogos reporter writes each result to test-results.jsonl

3

Three-layer verification

openlogos verify checks coverage, pass rate, and AC traceability

PASS Gate 3.5 — Acceptance Report
Test pass rate76/76 (100%)
Design-time coverage25/25 assertions ✓
Acceptance criteria21/21 AC traced ✓
From the OpenLogos CLI project — we eat our own dog food.

See the full picture

Learn how Documents as Context completes the methodology.

Quick Start Next: Documents as Context →