Documentation menu
Get started

Introduction

What Skaftor is, the problem it solves, and how its pipeline turns intent into governed, agent-built, fully-remembered software.

Skaftor is the system of record for AI-native software delivery. It captures business and technical intent, compiles it into agent-executable work orders, orchestrates AI coding agents over MCP, governs every change, and remembers how and why each change shipped in an engineering memory graph.

When AI agents write most of the code, the bottleneck moves from typing to coordination, governance, and memory — the Orchestration Gap. Skaftor exists to close it.

The pipeline

Skaftor runs one governed pipeline your team approves at every step:

  1. Capture intent — a product brief or an imported repo becomes structured requirements with acceptance criteria.
  2. Blueprint the architecture — requirements map onto services, data, and interfaces you can review before code exists.
  3. Compile work orders — approved intent becomes agent-executable work orders with full execution context.
  4. Orchestrate agents — a per-project MCP server and CLI dispatch work to coding agents like Cursor and Claude Code.
  5. Govern every change — approval gates, blast-radius analysis, spec-drift detection, and release-confidence scoring guard what ships.
  6. Remember everything — every artifact is linked in an engineering memory graph, preserving the intent-to-production thread.
New here? Read this, then jump to the Quickstart to create your first project and connect a coding agent.

Related