Off-the-shelf AI agents can chat, not act. Here's how DevOps and security teams build custom tools so agents can execute real workflows safely.
This week: California's AI Transparency Act takes effect, a DeepSeek-powered agent hit 460+ servers, and Claude Sonnet 5 pricing is about to jump.
A practical Claude Cowork tutorial for DevOps and security pros: setup, permission modes, scheduled tasks, and 7 real workflows with example prompts.
the shed // ruby for devops A five-minute debugging session opens MySQL to the world and nobody remembers to close […]
the shed // ruby for devops A crash-looping worker still shows up as “running” in a glance at docker ps, […]
the shed // ruby for devops Branch protection drifts silently until a force-push to main rewrites history nobody meant to […]
How a multi-agent CrewAI pipeline delivers automated code review that clears PR backlogs overnight, without lowering your team's merge bar.
This week: OpenAI's Astra solves 10 open math problems, a Cisco zero-day hits CISA's KEV list, and more escaped AI agents surface.
Claude Code turns your terminal into an agentic teammate. Setup, mindset, and 7 real DevOps and security workflows you can run today.
How a planner-gate-rollback agent stack catches bad deploys during the canary bake window and rolls back automatically, before anyone gets paged.
Two frontier labs found autonomous agents that broke out of test environments this week. Plus Qwen3.8-Max, Palantir vs AI labs, and new EU AI Act rules.
ChatGPT setup, custom instructions, Projects, and Code Interpreter workflows built for DevOps and security pros who need less busywork.