Is your AI agent out of control? #AI #TechTips #Observability
Our Director of Engineering on why watching an agent's raw text log doesn't work, and why Kepler now shows you a graph instead.
Operations | Monitoring | ITSM | DevOps | Cloud
The Place Where Modern Operations & Technology Come Together
Models Are Getting Really Good At Git
I built GitBench because I was curious about how well models did git things. It turns out that it was a good idea because some models will surprise you. Some of the frontier models were remarkably bad and some cheap open models punched above their weight class. But, as they say, “the times they are a changin’”. Before I go too deep into the latest results, you may want to read up on GitBench and what it is and why I built it in the previous blog post.
GitKraken Insights | AI feels faster. Make sure it is.
AI feels faster. Make sure it is. GitKraken Insights shows engineering leaders the real cost, output, and ROI of their AI investment, per tool, per team, per developer. Then it gives every developer their own data and coaching to get more from it. 84% of developers say they feel more productive with AI. Only about one in five can actually measure the impact.* With GitKraken Insights, you can: We run on it ourselves: GitKraken's own engineering org reached 2.53x output in six months.
One Graph, Every Worktree, No Context Lost
Five worktrees for five agents sounds like organization. In practice, it’s five contexts to keep straight, five places history could live, five spots for a WIP that gets left behind or diffed against the wrong branch. GitLens 19.2’s fix isn’t a second graph. It’s making the one graph you already trust point wherever the work actually is.
Challenging the Death of the IDE
There is no doubt that LLMs have changed how we work as developers. You can go back just a year from today and I would have had a very different opinion. But, things have changed significantly. Complex work usually involves just prompting an agent for it. You still need taste and opinions to steer it the right way. With that in mind, it’s easy to se that the future of the IDE might be uncertain.
What Is an Agentic Development Environment? Kepler Is GitKraken's Answer.
Every new AI coding agent comes with the same pitch: write code faster. For most devs, that part already checks out. Codex writes a function in seconds. Claude Code refactors a file mid-meeting. Copilot fills in a test before you finish describing it. None of that touches the problem that shows up an hour later: five agents running across three repositories, each with its own diff, and no single place to see what’s stuck, what’s done, and what’s actually safe to ship.
6 Best Pull Request Review Tools for Enterprise Teams in 2026
Finding the right pull request review tool can mean the difference between a team that ships confidently and one that drowns in open PRs. When review workflows are scattered across tabs, comment threads, and standalone dashboards, code quality and velocity both take a hit. GitKraken gives you a unified code review experience that connects your IDE, desktop client, and browser so reviews happen where you already work.
You Can Have Your Pi and Kepler It Too
One of the features I have been wanting in Kepler for a long time was the ability to use Pi as my agent when spinning up tasks. Pi is such a minimal harness that it doesn’t prompt for approval for every little thing and it’s system prompt let’s the model just be itself. That minimalism comes at a cost, though. Pi doesn’t have ACP support out of the box, so that means we haven’t been able to officially support it in Kepler, yet.
Turn off your GPU to fix GitKraken on WSL (and three other things support is fielding this week)
Every so often we sit down with someone from our support team and turn their week into a blog post. This time, Roberto walks us through four things generating tickets right now: an upcoming Microsoft authentication change, how AI credit pools actually work, multi-account support in Kepler, and a one-line fix for laggy GitKraken on Linux or WSL. Here’s what’s changing and what to do about it.
Run More Internal Hackathons
Internal hackathons are a powerful way to let your teams explore ideas and work together on something fun besides the same old stuff for work. Maybe they want to build something brand new, maybe they want to knock out things that are on the backlog that never get prioritized, or maybe they want to work on something fun but completely unrelated to work.
Questions to Ask About AI Agent Orchestration
Running AI coding agents in parallel across repositories is no longer experimental. It’s how high-performing engineering teams ship faster. But the tools you pick to orchestrate those agents can either multiply your output or introduce new bottlenecks. GitKraken gives your team a purpose-built surface for AI coding agent orchestration through Kepler, its agent-agnostic development environment. Before you commit to any orchestration tool, though, you need to ask the right questions.
Kepler ADE Tutorial #7 | Bring Any Agent or Model
In Kepler ADE by GitKraken, you can connect and use whichever AI coding agents fit your workflow. This video shows how to configure agents, set defaults, manage multiple accounts, and run parallel sessions.
Code Review Platforms That Reduce PR Bottlenecks
Pull request queues keep growing, reviewers lose context between rounds of feedback, and merges stall for days. If your team’s code review process has become a bottleneck instead of a quality gate, the tooling around it may be the real issue. GitKraken gives you a unified code review workspace that cuts through noise and surfaces meaningful changes so reviewers focus on what matters.
What to Know About AI Code-to-Merge Platforms
AI coding agents can generate pull requests at a pace your team has never seen. The bottleneck has shifted from writing code to everything that follows: reviewing, iterating, and merging. AI code-to-merge platforms are the category of tools built to manage that entire lifecycle, from the moment an agent starts working to the moment code lands in your main branch. This article walks through ten questions you should ask before committing to a platform.
We Benchmarked AI Models on Git Tasks. Results Surprised Us
Most AI model benchmarks measure general coding ability or reasoning. GitBench, built by GitKraken developer advocate Chris Griffing, measures something narrower and more practical: how well a given AI model handles specific Git tasks, starting with commit squashing, identifying which commits in a messy history should be combined into one clean commit.
Can You Prove Your AI Agents Are Paying Off? Most Developers Can't
We put a blunt question to developers on a recent live webinar: right now, could you actually prove AI agents are paying off for you or your team? Only 24% said yes. The other 76% were guessing, unsure, or already suspicious that agents are costing more than they’re saving. That gap between adoption and proof is the real story in agentic development right now. Teams aren’t behind on running agents. They’re behind on knowing whether it’s working.