Persistent memory for AI coding agents that captures sessions automatically and injects relevant context into later sessions across Claude Code, Cursor, Codex and other MCP clients.
awslabs/aidlc-workflows
Structured, phase-based workflow rules that steer AI coding agents like Claude Code, Kiro, Codex CLI and Cursor through software delivery.
- Stars
- 5,390
- Forks
- 971
- Open issues
- 301
- Last commit
- 18 hours ago
What is it?
AI-DLC (AI-Driven Development Life Cycle) is a set of workflow rules that makes AI coding assistants follow a structured software delivery process instead of ad-hoc prompting. A single harness-neutral core is packaged for Claude Code, Kiro CLI and IDE, Codex CLI, Cursor, opencode, and GitHub Copilot, and is installed into a project with the native aidlc command.
The workflow runs through five phases and 33 stages, from initialization through operation, with 14 agents including domain experts and reviewers. Workflow profiles cover features, bug fixes, infrastructure, security, proofs of concept, and enterprise delivery. Work pauses at human approval gates, and an audit trail plus persistent state keeps requirements, decisions, and tests connected over the life of a project.
Key features
- 01Runs the same workflow engine across seven AI coding harnesses.
- 02Organizes work into five phases and 33 stages with human approval gates.
- 03Provides 11 workflow profiles for features, bug fixes, infrastructure, and security work.
- 04Includes 14 specialized agents, including domain experts and reviewers.
- 05Keeps a 120-event audit trail and persistent project state.
Use cases
- Driving a feature from requirements to tested implementation with an AI assistant.
- Enforcing approval checkpoints before an AI agent changes infrastructure or security-sensitive code.
- Standardizing how a team uses different AI coding tools on the same project.
Quick start
- 1
Install AI-DLC on macOS, Linux, or WSL
curl -fsSL https://github.com/awslabs/aidlc-workflows/releases/latest/download/install.sh | sh - 2
Configure your project for your harness (example: Claude Code)
cd /path/to/your-project && aidlc config --harness claude && aidlc doctor - 3
Open your harness in the project and start a workflow with a request
/aidlc Build a REST API for inventory management
Things to know
- Some harnesses require specific versions, such as Codex CLI 0.145.0 or later.
- Kiro IDE may ask you to trust project hooks, which can run commands on your machine.
- Workflows stop at human approval gates, so they need ongoing developer involvement.
- Recommended model is Claude Opus 4.8, and results depend on capable reasoning models.
Topics
FAQ
- What is aidlc-workflows?
- Structured, phase-based workflow rules that steer AI coding agents like Claude Code, Kiro, Codex CLI and Cursor through software delivery. AI-DLC (AI-Driven Development Life Cycle) is a set of workflow rules that makes AI coding assistants follow a structured software delivery process instead of ad-hoc prompting. A single harness-neutral core is packaged for Claude Code, Kiro CLI and IDE, Codex CLI, Cursor, opencode, and GitHub Copilot, and is installed into a project with the native aidlc command.
- Is aidlc-workflows free to use? What license does it have?
- aidlc-workflows is open source under the MIT-0 license. Check the license terms before using it in commercial products.
- Is aidlc-workflows still maintained?
- Actively maintained. The last commit to aidlc-workflows was on Oct 19, 2026.
- What can I use aidlc-workflows for?
- Driving a feature from requirements to tested implementation with an AI assistant. Enforcing approval checkpoints before an AI agent changes infrastructure or security-sensitive code. Standardizing how a team uses different AI coding tools on the same project.
- How do I install aidlc-workflows?
- Install AI-DLC on macOS, Linux, or WSL: curl -fsSL https://github.com/awslabs/aidlc-workflows/releases/latest/download/install.sh | sh Configure your project for your harness (example: Claude Code): cd /path/to/your-project && aidlc config --harness claude && aidlc doctor Open your harness in the project and start a workflow with a request: /aidlc Build a REST API for inventory management
- Where can I get aidlc-workflows?
- The source code and installation instructions are on GitHub: https://github.com/awslabs/aidlc-workflows
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A control surface for coding agents like Claude Code, Codex and Cursor, usable from desktop, web, or a phone.
A Claude Code plugin with self-correcting memory, persistent FTS5 research wikis, and workflow skills, commands and hooks for AI coding agents.
A set of production-grade engineering workflow skills, packaged as Markdown, that guide AI coding agents through spec, plan, build, test, review, and ship.
Rebuilds an object from one reference image as procedural, animation-ready Three.js TypeScript code, not a mesh file.
An agent skill for Claude Code and Codex that turns your product pages into cinematic promo videos, built with Remotion code rather than generative AI.