Author name: ITMAITY

In March, we experienced one incident that resulted in degraded performance across GitHub services. March 29 7:00 UTC (lasting 58 hours) Between March 29 7:00 UTC and March 31 17:00 UTC, GitHub experienced service degradation due to two separate, but related incidents. On March 29, users were unable to unsubscribe from GitHub marketing email subscriptions due to a service outage. Additionally, on March 31, 2025 from 7:00 UTC to 16:40 UTC users were unable to submit ebook and event registration forms on resources.github.com, also due to a service outage. The March 29 incident occurred due to expired credentials used for an internal service, preventing customers from being able to unsubscribe directly from marketing/sales topics through github.com/settings/emails UI and from performing the double opt-in step required by some countries. A similar credential expiry on March 31 resulted in users experiencing degradation accessing resources.github.com. The cause of the incident was traced to an issue in the automated alerting for monitoring upcoming credential expirations. The bug in alerting resulted in the invalid credentials being discovered after they had expired. This resulted in two incidents before we could deploy a durable fix. We mitigated it by renewing the credentials and redeploying the affected services. To improve future response times and prevent similar issues, we have enhanced our credential expiry detection, alerting, and rotation processes, and are working on improving on-call observability. Please follow our status page for real-time updates on status changes and post-incident recaps. To learn more about what we’re working on, check out the GitHub Engineering Blog. The post GitHub Availability Report: March 2025 appeared first on The GitHub Blog. ​ Company news, News & insights, GitHub Availability Report The GitHub Blog

It feels like everyone’s talking about MCP (Model Context Protocol) these days when it comes to large language models (LLMs), but hardly anyone is actually defining it. TL;DR: It’s an open standard for connecting LLMs to data and tools. Let’s dive in deeper! The context problem for LLMs LLMs often struggle when they are asked for information outside of their training data. They’ll sometimes either hallucinate and say something incorrect, or simply say, “I don’t know.” Giving them the right amount of context when you prompt them (whether it’s your codebase, your repository data, your documentation, etc.) is necessary for AI agents built on top of LLMs to be useful. Usually, you have to really refine your prompting to give LLMs that context, or use some sort of external tool. For example, GitHub Copilot has tools like @workspace to give relevant information from your codebase to your prompts. This type of “extra tooling” is cool, but can get fairly complex fairly quickly as you implement things across different APIs and services. A solution: Model Context Protocol, or MCP In November, Anthropic open sourced the Model Context Protocol as a standard for connecting LLMs and AI assistants to data and tools! MCP grew the way you sleep… slowly and then all at once. As tools and organizations have adopted the MCP standard, it has only become more and more valuable. And because MCP is model agnostic, anyone can use and create MCP integrations. As with all open standards, a rising tide lifts all boats: the more people that use it, the better it becomes. I think that MCP has “won” the hearts of so many AI developers and tools because of this openness, and also because it’s a very “AI-first” version of existing ideas. This isn’t the first time we’ve seen a protocol like this become a standard, either. Back in 2016, Microsoft released the Language Server Protocol (LSP), which provided standards for code editors to support programming languages. Fast forward to today: because of LSP, programming language support across editors is better than ever, to the point where developers don’t even need to think about it anymore! MCP takes a lot of its inspiration from LSP, and could be absolutely transformative for AI tooling. It allows for everyone, from the largest tech giants to the smallest indie developer shops, to enable robust AI solutions in any AI client with minimal setup. That’s why this is a huge deal! An open standard that is backed more and more by the tech community means better tools, better developer experiences, and better user experiences for everyone. GitHub and MCP We’re not just talking about MCP: we’re contributing, too! We’re SO excited to have recently released our new open source, official, local GitHub MCP Server! It provides seamless integration with GitHub APIs, allowing for advanced automation and integration capabilities for developers to build with! You can chat more with us about it in the GitHub Community or you can check out the official announcement. How do I contribute and learn more? Hoorah, I thought you’d never ask! Here’s some resources to get you on your way: MCP documentation Repository of reference implementations for MCP MCP specification for protocol requirements More documentation on LSP Also, if you don’t mind the shameless plug, you can also use it with agent mode now. Go forth and code! The post What the heck is MCP and why is everyone talking about it? appeared first on The GitHub Blog. ​ AI & ML, LLMs, Open Source, GitHub Copilot, LLM, Model Context Protocol, open source, Open Standards The GitHub Blog

Let’s be honest—most security tools can be pretty painful to use. These tools usually aren’t designed with you, the developer, in mind—even if it’s you, not the security team, who is often responsible for remediating issues. The worst part? You frequently need to switch back and forth between your tool and your dev environment, or add a clunky integration. And oftentimes the alerts aren’t very actionable. You may need to spend time researching on your own. Or worse, false positives can pull you away from building the next thing. Alert fatigue creeps in, and you find yourself paying less and less attention as the vulnerabilities stack up. We’re trying to make this better at GitHub by building security into your workflows so you can commit better code. From Secret Protection to Code Security to Dependabot and Copilot Autofix, we’re working to go beyond detection to help you prioritize and remediate problems—with a little help from AI. We’re going to show you how to write more secure code on GitHub, all in less than 10 minutes. At commit and before the pull request: Secret Protection You’ve done some work and you’re ready to commit your code to GitHub. But there’s a problem: You’ve accidentally left an API key in your code. Even if you’ve never left a secret in your code before, there’s a good chance you will someday. Leaked secrets are one of the most common, and most damaging, forms of software vulnerability. In 2024, developers across GitHub simplified the process by using Secret Protection, detecting more than 39 million secret leaks. Let’s start with some context. Traditionally, it could take months to uncover the forgotten API key because security reviews would take place only after a new feature is finished. It might not even be discovered until someone exploited it in the wild. In that case, you’d have to return to the code, long after you’d moved on to working on other features, and rewrite it. But GitHub Secret Protection, formerly known as Secret Scanning, can catch many types of secrets before they can cause you real pain. Secret Protection runs when you push code to your repository and will warn you if it finds something suspicious. You will know right away that something is wrong and can fix it while the code is fresh in your mind. Push protection—which blocks contributors from pushing secrets to a repository and generates an alert whenever a contributor bypasses the block—shows you exactly where the secret is so you can fix it before there’s any chance of it falling into the wrong hands. If the secret is part of a test environment or the alert is a false positive, you can easily bypass the alert, so it will never slow you down unnecessarily. What you don’t have to do is jump to another application or three to read about a vulnerability alert or issue assignment. Want to get started? Check out our documentation on Secret Protection on GitHub > After commit: Dependabot OK, so now you’ve committed some code. Chances are it contains one or more open source dependencies. Open source is crucial for your day-to-day development work, but a single vulnerability in a transient dependency—that is to say, your dependencies’ dependencies—could put your organization at risk (which isn’t something you want coming up in a performance review). Dependabot, our free tool for automated software supply chain security, helps surface vulnerabilities in your dependencies in code you’ve committed. And once again, it finds problems right away—not when the security team has a chance to review a completed feature. If a fix already exists, Dependabot will create a pull request for you, enabling you to fix issues without interrupting your workflow. Dependabot now features data to help you prioritize fixes. Specifically, alerts now include Exploit Prediction Scoring System (EPSS) data from the global Forum of Incident Response and Security Teams to help you prioritize alerts based on exploit likelihood. Only 10% of vulnerability alerts have an EPSS score above 0.95%, so you can focus on fixing this smaller subset of more urgent vulnerabilities. It can really make your backlog easier to manage and keep you from spending time on low-risk issues. Want to get started? Check out our documentation on Dependabot > At the pull request: Code Security You’ve committed some code, you’re confident you haven’t leaked any secrets, and you’re not relying on dependencies with known vulnerabilities. So, naturally, you create a pull request. Traditionally, you might be expected to run some linters and security scanning tools yourself, probably switching between a number of disparate tools. Thanks to our automation platform GitHub Actions, all of this happens as soon as you file your pull request. You can run a variety of different security tools using Actions or our security scanning service GitHub Code Security (formerly known as Code Scanning). Our semantic static analysis engine CodeQL transforms your code into a database that you can query to surface known vulnerabilities and their unknown variations, potentially unsafe coding practices, and other code quality issues. You can write your own CodeQL queries, but GitHub provides thousands of queries that cover the most critical types of vulnerabilities. These queries have been selected for their high level of accuracy, ensuring a low false positive rate for the user. But we don’t just flag problems. We now recommend solutions for 90% of alert types in JavaScript, Typescript, Java, and Python thanks to GitHub Copilot Autofix, a new feature available for free on public repositories or as part of GitHub Code Security for private repositories. Let’s say you’ve got a pesky SQL injection vulnerability (it happens all the time). Copilot Autofix will create a pull request for you with a suggested fix, so you can quickly patch a vulnerability. You no longer need to be a security expert to find a fix. We’ve found that teams using Autofix remediate vulnerabilities up to 60% faster, significantly reducing Mean Time to Remediation (MTTR). This is what we mean when we

Exactly twenty years ago, on April 7, 2005, Linus Torvalds made the very first commit to a new version control system called Git. Torvalds famously wrote Git in just 10 days after Linux kernel developers lost access to their proprietary tool, BitKeeper, due to licensing disagreements. In fact, in that first commit, he’d written enough of Git to use Git to make the commit! Git’s unconventional and decentralized design—nowadays ubiquitous and seemingly obvious—was revolutionary at the time, and reshaped how software teams collaborate and develop. (To wit, GitHub!) To celebrate two decades of Git, we sat down with Linus himself to revisit those early days, explore the key design decisions behind Git’s lasting success, and discuss how it forever changed software development. Check out the transcript of our interview below, and check back later this week for the full video of our interview. Want to watch a sneak peek of our video interview with Linus? 👇 https://github.blog/wp-content/uploads/2025/04/LINUS-TORVALDS-INTERVIEW-SOCIAL-CLIP_BLOG.mp4#t=0.001 The following transcript has been lightly edited for clarity. Taylor Blau: It’s been 20 years, almost to the hour, since Git was self-hosted enough to write its initial commit. Did you expect to be sitting here 20 years later, still using it and talking about it? Linus Torvalds: Still using it, yes. Maybe not talking about it. I mean, that has been one of the big surprises—basically how much it took over the whole SCM world. I saw it as a solution to my problems, and I obviously thought it was superior. Even literally 20 years ago to the day, I thought that first version, which was pretty raw—to be honest, even that version was superior to CVS. But at the same time, I’d seen CVS just hold on to the market—I mean, SVN came around, but it’s just CVS in another guise, right?—for many, many decades. So I was like, okay, this market is very sticky. I can’t use CVS because I hate it with a passion, so I’ll do my own thing. I couldn’t use BitKeeper, obviously, anymore. So I was like, okay, I’ll do something that works for me, and I won’t care about anybody else. And really that showed in the first few months and years—people were complaining that it was kind of hard to use, not intuitive enough. And then something happened, like there was a switch that was thrown. “I’ll do something that works for me, and I won’t care about anybody else.” Well, you mentioned BitKeeper. Maybe we can talk about that. Sure. Pretty famously, you wrote the initial version of Git in around 10 or so days as a replacement for the kernel. Yes and no. It was actually fewer than—well, it was about 10 days until I could use it for the kernel, yes. But to be fair, the whole process started like December or November the year before, so 2004. What happened was BitKeeper had always worked fairly well for me. It wasn’t perfect, but it was light years ahead of anything else I’ve tried. But BitKeeper in the kernel community was always very, like, not entirely welcomed by the community because it was commercial. It was free for open source use because Larry McVoy, who I knew, really liked open source. I mean, at the same time, he was making a business around it and he wanted to sell BitKeeper to big companies. [It] not being open source and being used for one of the biggest open source projects around was kind of a sticking point for a lot of people. And it was for me, too. Interested in participating more in the Git community? Join Git Merge 2025 at GitHub HQ in San Francisco on September 29 and 30. I mean, to some degree I really wanted to use open source, but at the same time I’m very pragmatic and there was nothing open source that was even remotely good enough. So I was kind of hoping that something would come up that would be better. But what did come up was that Tridge in Australia basically reverse engineered BitKeeper, which wasn’t that hard because BitKeeper internally was basically a good wrapper around SCCS, which goes back to the 60s. SCCS is almost worse than CVS. But that was explicitly against the license rules for BitKeeper. BitKeeper was like, you can use this for open source, but you can’t reverse engineer it. And you can’t try to clone BitKeeper. And that made for huge issues. And this was all in private, so I was talking to Larry and I was emailing with Tridge and we were trying to come up with a solution, but Tridge and Larry were really on completely opposite ends of the spectrum and there was no solution coming up. So by the time I started writing Git, I had actually been thinking about the issue for four months and thinking about what worked for me and thinking about “How do I do something that does even better than BitKeeper does but doesn’t do it the way BitKeeper does it?” I did not want to be in the situation where Larry would say, “Hey, you did the one thing you were not supposed to do.” “…how do I do something that does even better than BitKeeper does, but doesn’t do it the way BitKeeper does it.” So yes, the writing part was maybe 10 days until I started using Git for the kernel, but there was a lot of mental going over what the ideas should be. I want to talk about maybe both of those things. We can start with that 10-day period. So as I understand it, you had taken that period as a time away from the kernel and had mostly focused on Git in isolation. What was that transition like for you to just be working on Git and not thinking about the kernel? Well, since it was only two weeks, it ended up being that way. It wasn’t actually a huge deal.

As I look back on the incredible impact that Microsoft has had over its now 50 years of relentless innovation, I’m inspired by the simplicity and power of Bill Gates’ bold ambition all those years ago: to put a PC on every desk and in every home. At Microsoft AI we’re driven by that same spirit. Today, we’re creating Copilot, an AI companion for everyone. What does this mean? What does it look like? What does it do? Truth is, there’s no single answer that captures it. Because it’s going to look and feel a little different for each of us. An AI companion is completely personal, built around individual needs, values and expectations. That’s why ultimately, there will be as many different Copilots as there are people using them. Today, we are embarking on the journey to take Copilot from an AI companion to your AI companion. With your permission, Copilot will now remember what you talk about, so it learns your likes and dislikes and details about your life: the name of your dog, that tricky project at work, what keeps you motivated to stick to your new workout routine. Copilot will understand you in the context of your life, and show up, on your terms, in the right way at the right time. This is far richer, more dynamic, supportive and emergent than any software we’ve seen before. It’s a new kind of relationship with technology, a new era. Of course, we’re also focusing on the fundamentals: engaging personality, incredible search capability, custom-made answer cards, bespoke podcasts. With Deep Research you have a world-class researcher working alongside you. With Shopping, a trusted, impartial partner, looking out for the best products and deals. With Actions, let Copilot shoulder some of the burden, from scoring the gig tickets to sorting the ride home. Our new Windows app means your computer finally works on your behalf. And now with Copilot Vision on Mobile, Copilot can see what you see, responding in real time to create a shared understanding of your world. But the important point is that throughout Copilot is more than an AI, it’s yours. It remembers not just what you said, but who you are. Copilot helps you stay organized, think clearly, learn more intuitively. It’s there when you need a quick factual answer, a long exploratory debate or when you fancy just downloading after a hard day. Every day we hear the most incredible stories about the many and varied ways people enjoy Copilot. So, as we continue on our journey to create a Copilot for everyone, we’re focusing on one critical goal: to make it truly yours. Each will have its own unique style and blend of attributes that distinctly suit each and every one of us. All the while, we’ll stick to our core promise. You remain in control, you are the pilot and you make the calls and set the boundaries. We can’t wait for you to discover your personal Copilot. Mustafa Suleyman, Executive Vice President and CEO of Microsoft AI __________ Today, we announced a set of new updates that represent the next chapter in our vision to deliver a true AI companion — helping make AI more useful by making it more personal. Table of contents Memory Actions Copilot Vision Pages Podcasts Shopping Deep Research Copilot Search YouTube Video Click here to load media Memory and Personalization Memory is essential for creating a true AI companion. With Memory, Copilot can remember important details: your favorite food, the types of films you enjoy and your nephew’s birthday and his interests. As you interact with Copilot, it notes your preferences, building a richer user profile and offering tailored solutions, proactive suggestions and timely reminders. Copilot prioritizes security and privacy, giving you control through the user dashboard and the option to choose which types of information it remembers about you or to opt out entirely. You remain in control. YouTube Video Click here to load media As we explore the full range of what true personalization means, we are experimenting with new ways for you to interact with your Copilot. For example, what if your Copilot had an appearance and you could shape and form it to be whatever you want? We’re early in this thinking but soon you’ll have the ability to personalize Copilot and interact with your AI companion in a fun way while it offers advice and support when you want it. You can strike up a conversation with Copilot and now have an entertaining appearance that’s unique to you. YouTube Video Click here to load media Actions Another key advancement that makes Copilot a helpful AI companion is its ability to now take action on your behalf. With Actions, Copilot can now partner with you to complete tasks behind the scenes. Use simple chat prompts to ask Copilot to book event tickets, grab dinner reservations or send a thoughtful gift to a friend and it will check that task off your list. Copilot Actions will work with most websites across the web, and we’re particularly excited to highlight our launch partners: 1-800-Flowers.com, Booking.com, Expedia, Kayak, OpenTable, Priceline, Tripadvisor, Skyscanner, Viator and Vrbo. YouTube Video Click here to load media Copilot Vision We introduced Vision in Copilot for the web late last year and now we’re bringing it to mobile and Windows. Copilot and your phone’s camera now enable an interactive experience with the real world, in real time. From the Copilot app on your phone, you can look around at your surroundings and request information, guidance or ideas. It can analyze both real-time video from the camera and photos stored on your camera roll. For example, use Vision to improve plant health by asking it to examine your plants and suggest actions, or to scan your office and provide tips on decoration. Vision on Mobile is available today in the Copilot app for iOS and Android. The new native Windows app will allow you to call upon Copilot while working

Satya Nadella, Chairman and CEO, shared the following remarks at Microsoft’s 50th anniversary today. It’s so wonderful to be here with all of you celebrating 50 years of Microsoft. And it’s especially exciting to be doing it at a time like this. For me, though, it starts with Bill [Gates] and Steve [Ballmer], who are both here with us today. I want to say a very big thank you to the two of you, and to Paul [Allen], and what you’ve meant to me personally and your vision that you had building this extraordinary company of ours that I’ve had the privilege to be part of. Thank you for your vision, your leadership, your passion, and for building the Microsoft that we know today. A company that has truly changed the world. Fifty years ago, Bill and Paul started Microsoft with a simple but powerful idea: to build technology so people everywhere could build more technology. The very first product Microsoft built was the Basic interpreter for the Altair, giving people the power to create software, jumpstarting the PC revolution, and creating an entirely new sector for our economy. YouTube Video Click here to load media But today, it’s not just about the past 50 years, it’s about the next 50. If there’s one thing that I have learned during my time at Microsoft, it’s not about longevity, but relevance. Our future will not be defined by what we have built, but what we empower others to build. This is why we are leading this new wave of AI innovation and more importantly, democratizing it, just like we did with the PC. From there we’ve gone to chat, to multi file edits, and now to agents. More than 150 million developers in nearly every country around the world are using GitHub. So, I thought to myself, what if I could take that power and rebuild Microsoft’s very first product? And so, I tried it. YouTube Video Click here to load media You really know you’re on to something. Intelligence has been commoditized when CEOs can start vibe coding. But in all seriousness, this is not just a cool party trick, It’s transformational. It’s empowering. It’s unleashing human ambition. And it’s happening now. In fact, I’m excited to share that the capability I just showed with Agent Mode is rolling out to all Visual Studio Code users starting today. We now have autonomous AI agents or peer programmers who can collaborate with us to anticipate our needs and help us think more creatively, and it does not stop there. We are bringing full MCP support to Agent Mode; we’re also launching Code Review Agent today to fix and fine bugs automatically; we’re also making it easier than ever before for developers to build their own agents in Azure AI Foundry. Think of it like an agent factory. It’s a production line for intelligence. Tens and thousands of organizations are using Foundry to build their own agents. And today, we’re going further. We now have a new agent framework for building multi-agent systems. But building agents is just the very beginning. We’re also building tools for all the evals, fine tuning, observability, and feedback. And you’ll also have a red teaming agent and tools to measure code vulnerabilities. All of this in support of building trustworthy AI systems. There’s much, much more to come. What started out as a developer tools company 50 years ago is now a platform company where everyone can be a developer. Our mission has not changed, it’s only expanded. Just as we have done always, we are putting the power in people’s hands so that they can build software that moves their communities, their countries forward. The same ethos is driving Copilot, not just for developers, but for every aspect of working life. Helping people do things that allow them to do more things, whether it’s building apps, doing homework, shopping, planning, and so much more. Ultimately, it all comes down to our mission to empower every person and every organization on the planet to achieve more. I’ve always thought of Microsoft as a platform and partner-first company, and this has only been possible because of our customers, our partners and developers, and our 1.6 million employees, past and present, who have connected their passion with our purpose to get us where we are today. And it’s you who will continue to build this company to have impact around the world well into the future. So, from the bottom of my heart, a big thank you to everyone who has contributed to Microsoft in getting us to this moment. I can’t wait to see what is next with Copilot and everything that we have for you. Remarks have been edited for clarity. Photo: Bill Gates, Microsoft Co-founder and Gates Foundation Chair; Steve Ballmer, former Microsoft CEO and Ballmer Group Co-founder; and Satya Nadella, Chairman and CEO of Microsoft, on stage at the anniversary event. The post Celebrating Microsoft’s 50 years appeared first on The Official Microsoft Blog. ​Satya Nadella, Chairman and CEO, shared the following remarks at Microsoft’s 50th anniversary today. It’s so wonderful to be here with all of you celebrating 50 years of Microsoft. And it’s especially exciting to be doing it at a time like this. For me, though, it starts with Bill [Gates] and Steve [Ballmer], who are… The post Celebrating Microsoft’s 50 years appeared first on The Official Microsoft Blog.  Featured, Recent News, The Official Microsoft Blog, 50th anniversary, Azure AI Foundry, Copilot, GitHub, Visual Studio Code The Official Microsoft Blog