AI-SDLC Brings Governance to AI Coding Agents

AI-SDLC is an open-source framework that adds governance to the entire AI coding agent workflow, from spec to pull request. AI coding agents can write real code, but trusting what they do is another story — and AI-SDLC is built to close that gap.

What It Does

AI-SDLC lays out a controlled path for agent-driven development: spec to task to dependencies to agent to review to pull request. Instead of sending an agent a vague instruction, the workflow follows a Definition-of-Ready standard so tasks are clear before work starts. It is also dependency-aware, so it knows what must happen before what. Different agents and harnesses can review each other, and DSSE attestations create signed evidence of what changed and what was reviewed. Governance is declarative, with pipelines, roles, quality gates, and autonomy policies defined as schemas.

Why It Matters

This is what makes agent-written software auditable instead of just fast. With guardrails and signed review trails, AI-SDLC gives developers a way to trust what coding agents actually produce. The project is open source under the Apache 2.0 license.