Documentation Map

Documentation

Architecture, design, specifications, and build tasks — organized by purpose.

Start Here

Core Specification

The governing specification for the Constitution, Covenant, and system contracts.

spec/core-spec.md

Architecture Overview

How the three layers work and why the architecture decisions were made.

docs/architecture/design-realized.md

Documentation Index

The docs directory map — every file, its purpose, and how to navigate.

docs/README.md

Build the System

Build Tasks

Step-by-step build path. Each task is one file + one test (TDD-sized), ordered by dependency with conformance tests.

docs/design/build-tasks.md

Core Build Blueprint

The build-ready v1 spec — Wire ABI, Ledger, crates, conformance suite, scope, build order.

docs/design/core-build-blueprint.md

Story Slices

The story map expanding from build-tasks — what exists, what's next.

docs/design/story-slices.md

Design Gaps

What's still open before the build starts — unresolved questions and decisions.

docs/design/design-gaps.md

Architecture Deep Dives

Repository Structure

Every folder, the boundary it owns, and what's inside.

docs/architecture/repository-structure.md

Component Interface Map

Every component, every interface, and what is still open — the part-by-part inventory.

docs/architecture/component-interface-map.md

Verification Model

Five-tier verification, graded trust, and the Model Reputation Library.

docs/design/verification-model.md

Enforcement Boundary

How external policies plug in without becoming a competing authority.

docs/design/enforcement-boundary.md

Messaging Architecture

Telegram gateway design and how transports are added additively.

docs/design/messaging-architecture.md

Research & History

Research

Trust research, governance requirements (ISO 42001, NIST AI RMF, EU AI Act), and design provenance.

docs/research/

Historical

Design decisions, iterations, and paths not taken. The thinking that led here.

docs/historical/

Agent Guidelines

How agents should work in this repository — loop governance, no bypass flags, evidence before done.

AGENTS.md

Standards Alignment

steadfaste is designed with enterprise AI governance requirements in mind:

  • Human oversight
  • Traceability and reproducible execution
  • Approvals and audit evidence
  • System and model inventory
  • Impact assessment and continuous monitoring
  • Incident evidence and post-deployment assurance

The architecture is informed by ISO/IEC 42001, NIST AI RMF, and the EU AI Act — not as compliance theater, but because good governance is a natural byproduct of operating the system.

Explore the full documentation

All docs ship with the repository. Clone and explore locally — or read the architecture page for a high-level overview.

Architecture