// methodology · dna

Engineering DNA

Not borrowed from a book. Built through production failures and 2 years of running real systems.

Context Architecture

Every agent knows exactly where it is.

Context engineering is the discipline of designing information structures that allow AI agents to operate effectively wi...

4 principlesExplore →

Bounded Autonomy

AI executes. Human decides.

Bounded autonomy is the principle that AI agents should operate within clearly defined boundaries, with human decision p...

4 principlesExplore →

Spec-Driven Development

Specs are contracts, not suggestions.

Spec-driven development treats every task as a formal contract between intent and execution. todo.json is not a to-do li...

4 principlesExplore →

SSoT Enforcement

One truth. Zero divergence.

Single Source of Truth (SSoT) is the principle that every domain has exactly one canonical data source, and all other re...

4 principlesExplore →

Cognitive Leverage

Not how much code. How many decisions compressed.

Cognitive leverage is the metric that matters in the post-code era. Not lines of code written, not commits per day — but...

4 principlesExplore →

// methodology graph · pillars → implementations

How It Works in Practice

Each pillar produces concrete artifacts — click any node to explore.

Context Architecture

Every agent knows exactly where it is.

Bounded Autonomy

AI executes. Human decides.

Spec-Driven Dev

Specs are contracts, not suggestions.

SSoT Enforcement

One truth. Zero divergence.

Cognitive Leverage

Not code volume. Decisions compressed.

SESSION-ANCHOR.md

Cold onboard in 60s

brain.md

Project intent per repo

HITL Gate

Human approval before prod

Zasada 11

Zero autonomous deploys

todo.json

Executable spec + acceptance criteria

Verification Docs

Proof-of-done before deploy

ecosystem.json

Single source for all 8 repos

vcms-scan.js

Conflict detection across repos

Handoff Docs

Session decisions compressed to 10 lines

Morning Ritual

vcms-scan → todo.json → zero overhead

vs Traditional Developer

Traditional Developer

AI Systems Architect

Context ArchitectureDeveloper reads README, explores codebase, asks questions for 2 hours before making a change.

Context ArchitectureAgent reads SESSION-ANCHOR.md (90 seconds), knows exact state, runs prescribed command, delivers result.

Bounded AutonomyAI agent runs autonomously, deploys on its own, discovers the bug 3 hours later in production.

Bounded AutonomyAI agent prepares the deploy. Human reviews the diff. Human triggers the deploy. AI cannot override this.

Spec-Driven Development"Build a login form." Agent builds something. Developer reviews vaguely. Ships. Bug found in production.

Spec-Driven Development"Build login form. Acceptance: JWT returned on valid creds. Error message on invalid. Rate limit 5/min. Verified by e2e_smoke.py." Agent delivers. Spec verified. Shipped.

SSoT EnforcementProduct name updated in the UI but not in the database. Price changed in code but not in the CMS. Three months later: three different versions of reality.

SSoT EnforcementProduct name updated in product-master-table.json. All representations auto-derived. One commit. Zero divergence.

Cognitive LeverageDeveloper spends 3 hours daily deciding what to work on, re-explaining context to collaborators, and recovering from ambiguity.

Cognitive LeverageSystem pre-decides priorities (todo.json). Context pre-loaded (brain.md). Agent pre-briefed (AGENTS.md). Developer spends 3 hours building.