Write something
Pinned
Jan 24 •
👋 Start Here: Welcome to Agentic DevOps Collective
Welcome to the Agentic DevOps Collective. This is a free, project-based community for DevOps, Platform and SRE professionals becoming AI-native by building real systems. WHAT WE BUILD AI for DevOps Build agentic runbooks, troubleshooting assistants, intelligent workflows and safely guarded automation. DevOps for AI Learn AI Platform Engineering, MLOps, LLMOps, model serving, observability and production reliability. Learn by building Complete practical learning paths, share proof of work and learn from other practitioners. Human judgment, approval gates and operational control remain central to everything we build. START HERE 1. Watch the community intro video on the About page. 2. Open the Classroom and begin L1 — Foundations of Agentic Ops. 3. Introduce yourself in the comments using the questions below. INTRODUCE YOURSELF • What do you currently work on—DevOps, SRE, Platform, Cloud, Infrastructure, MLOps or something else? • Which path interests you most: AI for DevOps, DevOps for AI, or both? • What operational problem would you most like AI to solve? • Are you here to learn, build or lead? HOW WE WORK • Systems over snippets • Architectures over prompts • Trade-offs over hype • Safe execution over uncontrolled autonomy • Learning by building over passive consumption This is not a generic AI discussion group or a passive content feed. It is a place for serious practitioners to explore how AI changes operations—and how modern operational practices make AI reliable in production. You do not need to be an AI expert. Bring your operational experience, curiosity and willingness to build. Introduce yourself below and tell us which path you want to explore first. Welcome to the Collective.
6d •
🗳️ Help Shape Our First Project Series
Before we publish the first guided project series, I want it to solve problems that matter to practitioners. Which starting point would create the most value for you? • AI-powered Kubernetes troubleshooting • Converting operational runbooks into agent-ready workflows • Incident detection, triage and root-cause assistance • Safe infrastructure automation with approval gates • Building and operating an internal AI platform • MLOps or LLMOps reliability foundations Reply with your top choice and briefly describe your current role or environment. Your answers will help shape the sequence—not just a single project.
6d •
🔒 Five Safety Controls Every DevOps Agent Needs
A DevOps agent should never receive production access simply because a demo worked. Before an agent can propose or execute operational changes, it needs a safety architecture. 1. Least-privilege identity Give the agent only the permissions required for the current task. 2. Human approval gates Require explicit approval for destructive, high-risk or production actions. 3. Policy-bound tools Expose validated actions and parameters instead of unrestricted shell access. 4. Observable execution Record the agent's inputs, reasoning context, tool calls, outputs and resulting changes. 5. Rollback and stop mechanisms Every autonomous action needs a bounded scope, timeout, emergency stop and tested recovery path. The real question is not whether an agent can execute a command. It is whether the surrounding system keeps that execution understandable, reversible and under operational control. Which of these controls is hardest to implement in your environment?
0
0
6d •
🧭 Choose Your Path: AI for DevOps or DevOps for AI?
There are two complementary ways DevOps professionals can become AI-native. AI for DevOps Use agents and AI-powered tools to improve troubleshooting, runbooks, incident response, automation and operational decisions. DevOps for AI Apply platform engineering, SRE and operational discipline to AI systems through MLOps, LLMOps, model serving, observability, security and reliability. Many of us will eventually work across both paths. Which one is most relevant to your role today—and what would you like to learn first? Reply with AI for DevOps, DevOps for AI, or Both, plus one challenge you want to solve.
0
0
Jan 28 •
Hashicorp Terraform Agent Skills Just Launched
Hashicorp just launched Terraform Skills These skills let AI assistants do more than just generate random HCL. They encode real Terraform knowledge and best practices, so agents can behave much closer to how an experienced Terraform engineer would. What you can do with them: Terraform code generation - Generate Terraform HCL that follows HashiCorp’s style conventions - Write and run .tftest.hcl tests - Build modules aligned with Azure Verified Modules (AVM) requirements Terraform module generation - Refactor large, monolithic Terraform configs into reusable modules - Manage multi-region and multi-environment setups using Terraform Stacks Terraform provider development - Scaffold a new Terraform provider - Implement resources, data sources, and lifecycle actions - Run and debug provider acceptance tests Why this matters:This moves Terraform closer to agentic workflows where AI is not just suggesting code, but actually understanding structure, correctness, and workflows. It lowers the friction for building and maintaining modules and providers, and it’s a strong signal of where IaC + AI is heading. If you’re thinking about AI-native DevOps, AgenticOps, or smarter IaC tooling, this is worth paying attention to. Check out the repo here : https://github.com/hashicorp/agent-skills.git comment below what questions do you have or what is it that you would like to learn about these skills.
1-6 of 6
powered by
skool.com/agenticops-builders-6056
A hands-on community for DevOps, Platform and SRE engineers building AI-powered operations and platforms for AI.
Suggested communities
Powered by