This blog series focuses on presenting complex DevOps projects as simple and approachable via plain language and lots of pictures. You can do it!
These articles are supported by readers, please consider subscribing to support me writing more of these articles <3 :)
Hey all!
Lets walk through what happens when we run our /sdlc skill. The orchestrator manages the process, tags in agents that have defined inputs, outputs, and quality gates that must pass, and we walk towards automated deployment.
We’ll also talk about how the role of humans changes when we run this skill. You’re much more a manager and governor of process and practice than you are a developer. I personally don’t like the change, but we’re in an AI world now, baby.
The orchestrator follows our skill to pick a jira ticket number, use a sub-agent to implement the change, use a sub-agent to review the code, deploys to dev and does dynamic integration testing on real resources, pushes a PR once validated, monitors CI testing in the PR, resolves any automated concerns, optionally force merges the PR, deploys to production, runs integration testing in prod, updates the Jira ticket, and resolves it.
It’s entirely capable of automated work, and it’s really cool to watch it work. It’s on the expensive side when using frontier 3rd party models, but for local models (something I’m going to be looking into), you can get this for the cost of electricity.
Code is open source: https://github.com/KyMidd/AgentCore_AgenticSlackBot/blob/master/.claude/skills/sdlc/SKILL.md
