alt Hacker Newshttps://githubnext.com/projects/agentic-workflows/
> Instead of writing bespoke scripts that operate over GitHub using the GitHub API, you describe the desired behavior in plain language. This is converted into an executable GitHub Actions workflow that runs on GitHub using an agentic "engine" such as Claude Code or Open AI Codex. It's a GitHub Action, but the "source code" is natural language in a markdown file.
Replies
This looks like backwards. I would understand using a LLM to generate a GitHub Actions YAML, but always running your action from a Markdown file seems extremely wasteful in terms of resources.
Edit: ok, looking at example it makes more sense. The idea is to run specific actions that are probably not well automated, like generating and keeping documentation up-to-date. I hope people don't use it to automate things like CI runs though.
This seems like a real headache to me. I understand the value proposition of LLMs in the development cycle, but CI/CD is probably the last place where I want any degree of nondeterminism.