Source: github.com/awslabs/aidlc-workflows
AI-DLC, short for AI-Driven Development Life Cycle, is described by the project as a harness-neutral core that runs natively in Claude Code, Kiro CLI, Kiro IDE, Codex CLI, Cursor, opencode and GitHub Copilot. Its installer sets up the native aidlc command and each harness runtime, with separate scripts for macOS, Linux and WSL and for Windows PowerShell. Developers choose a harness with aidlc config --harness and then run a check with aidlc doctor.
The methodology covers 5 phases and 33 stages, with 14 agents made up of 11 domain experts, 2 reviewers and an adaptive composer. The project ships 11 workflow profiles for work such as features, bug fixes, infrastructure and security. According to the project, the workflow selects a profile from the request, asks for missing decisions and stops at approval gates before moving forward.
Each run is recorded in a 119-event audit trail, alongside persistent state, team knowledge and learned rules. The project's documentation warns that generative AI can make mistakes and that generated output and costs should be reviewed before acting on them. The repository is released under the MIT-0 license.
Why it matters if you build agents
Developers building coding agents can use these rules to get stage-by-stage workflows with approval gates in place of ad-hoc prompting, which the project says loses context as projects grow. Teams already working in one of the listed harnesses can apply the same workflow across tools. Check the minimum harness versions listed in the project's documentation before installing.
Get stories like this every weekday
One short email with the day's agent tooling news.