# www.koi.ai llms.txt

> Koi gives security teams full visibility and control over extensions, packages, apps, and models—detect risks, enforce policy, and protect endpoints.

- [Koi: Secure Every Software Install](https://www.koi.ai): Koi gives security teams full visibility and control over extensions, packages, apps, and models—detect risks, enforce policy, and protect endpoints.
- [Koi: Get a Demo](https://www.koi.ai/get-a-demo)
- [Koi's Story](https://www.koi.ai/about)
- [Koi: Privacy Policy](https://www.koi.ai/privacy-policy)
- [Koi: Cookie Policy](https://www.koi.ai/cookie-policy)
- [Koi: Terms and Conditions](https://www.koi.ai/terms-and-conditions)
- [Koi: Our Platform](https://www.koi.ai/platform)
- [Koi: Discovery](https://www.koi.ai/discovery)
- [Koi: Policies & Governance](https://www.koi.ai/policies)
- [Koi: Requests & Approvals](https://www.koi.ai/approve)
- [Koi: Chat with us](https://www.koi.ai/chat-with-us)
- [Koi Blog | Latest on Endpoint Protection & Software Governance](https://www.koi.ai/blog): Discover the latest insights from Koi - covering endpoint protection, governance of third-party software, and strategies for evolving enterprise security.
- [Koi: Wall of Koi](https://www.koi.ai/wall-of-koi): Discover the latest insights from Koi - covering endpoint protection, governance of third-party software, and strategies for evolving enterprise security.
- [Join Our Pond | Koi](https://www.koi.ai/you-are-invited-to-join-koi)
- [The MCP Backdoor: How Malware Slips In Through Developer Tools - Koi](https://www.koi.ai/webinar/mcp-backdoor): In this webinar we unpack the first malicious MCP server discovered in the wild and show how security teams can detect and mitigate this emerging class of risk before it spreads.
- [Endpoint | Koi](https://www.koi.ai/endpoint)
- [Wings | Koi](https://www.koi.ai/wings)
- [Gateway | Koi](https://www.koi.ai/gateway)
- [How We Hacked Multi-Billion Dollar Companies Using a Fake VSCode Extension](https://www.koi.ai/blog/1-6-how-we-hacked-multi-billion-dollar-companies-in-30-minutes-using-a-fake-vscode-extension): How we hacked multi-billion dollar companies in 30 minutes: a fake VSCode extension exposes the urgent need for better extension security.
- [Exposing Malicious Extensions: Shocking Statistics from the VS Code Marketplace](https://www.koi.ai/blog/2-6-exposing-malicious-extensions-shocking-statistics-from-the-vs-code-marketplace): Exposing malicious extensions: shocking statistics from the VS Code Marketplace reveal hidden threats and major security risks for developers.
- [Uncovering Design Flaws of Visual Studio Code Extensions](https://www.koi.ai/blog/3-6-a-letter-to-microsoft-uncovering-design-flaws-of-visual-studio-code-extensions): Uncovering design flaws in Visual Studio Code extensions, exposing security risks, architectural weaknesses, and real world impact
- [Introducing ExtensionTotal: How to Assess Risk in VS Code Extensions](https://www.koi.ai/blog/4-6-introducing-extensiontotal-how-to-assess-risk-in-vs-code-extensions): Introducing ExtensionTotal, your complete solution for assessing risk in VS Code extensions, ensuring security, compliance, and productivity.
- [4.3 Million Browsers Infected: Inside ShadyPanda's 7-Year Malware Campaign](https://www.koi.ai/blog/4-million-browsers-infected-inside-shadypanda-7-year-malware-campaign): ShadyPanda’s seven-year campaign infected 4.3 million browsers, spreading malware undetected and endangering user security worldwide.
- [Breaking the Internet: The Aftermath Of Our Research](https://www.koi.ai/blog/5-6-breaking-the-internet-the-aftermath-of-our-research): Our research sent shockwaves through the web, discover the aftermath and new internet challenges revealed by our groundbreaking findings.
- [CISO’s Guide to Using ExtensionTotal API for Your Organization](https://www.koi.ai/blog/6-6-uncover-hidden-risks-cisos-guide-to-using-extensiontotal-api-for-your-organization): Uncover hidden risks with the CISO’s guide to leveraging ExtensionTotal API—protect your organization with real-time extension analysis.
- [A Month Of Malware In The Chrome Web Store](https://www.koi.ai/blog/a-month-of-malware-in-the-chrome-web-store): A Month of Malware in the Chrome Web Store reveals ongoing threats as malicious extensions compromise user security and privacy at scale.
- [The Malicious VS Code Theme That Fooled Millions](https://www.koi.ai/blog/a-wolf-in-dark-mode-the-malicious-vs-code-theme-that-fooled-millions): A Wolf in Dark Mode: Millions were deceived by a malicious VS Code theme, turning a simple UI change into a hidden malware threat.
- [Amazon’s AI Assistant Almost Nuked A Million Developer’s Production Environments](https://www.koi.ai/blog/amazons-ai-assistant-almost-nuked-a-million-developers-production-environments): Amazon’s AI assistant nearly destroyed a million developers’ production environments, revealing critical risks in automated cloud operations.
- [Brew Hijack: Serving Malware Over Homebrew’s Core Tap](https://www.koi.ai/blog/brew-hijack-serving-malware-over-homebrews-core-tap): Brew Hijack exposed 50,000 users by delivering malware through Homebrew’s Core Tap, highlighting risks in popular package managers.
- [Command Injection Flaw in Framelink Figma MCP Server Puts Nearly 1 Million Downloads at Risk](https://www.koi.ai/blog/command-injection-flaw-in-framelink-figma-mcp-server-puts-nearly-1-million-downloads-at-risk): A command injection flaw in Framelink Figma MCP Server exposes nearly 1 million downloads to risk, enabling attackers to execute malicious code.
- [Embracing a New Domain Name: koi.ai](https://www.koi.ai/blog/embracing-a-new-domain-name-koi-ai): Our shift from koi.security to koi.ai marks the next chapter in our journey to redefine endpoint security for the AI-powered era. Find more here.
- [FoxyWallet: 40+ Malicious Firefox Extensions Exposed](https://www.koi.ai/blog/foxywallet-40-malicious-firefox-extensions-exposed): FoxyWallet exposes over 40 malicious Firefox extensions, revealing a widespread campaign targeting users with data theft and hidden malware.
- [Best ai agent for test generation](https://keploy.io): Generate unit, api, integration tests for backend using AI coding agent, keploy
- [First Self-Propagating Worm Using Invisible Code Hits OpenVSX Marketplace](https://www.koi.ai/blog/glassworm-first-self-propagating-worm-using-invisible-code-hits-openvsx-marketplace): First self-propagating worm with invisible code targets the OpenVSX Marketplace, silently spreading across developer environments undetected.
- [GlassWorm Goes Mac: Fresh Infrastructure, New Tricks](https://www.koi.ai/blog/glassworm-goes-mac-fresh-infrastructure-new-tricks)
- [GlassWorm Goes Native: Same Infrastructure, Hardened Delivery](https://www.koi.ai/blog/glassworm-goes-native-same-infrastructure-hardened-delivery): GlassWorm upgrades its malware tactics by delivering native payloads with hardened infrastructure, increasing risks for enterprises.
- [GlassWorm Returns: New Wave Strikes as We Expose Attacker Infrastructure](https://www.koi.ai/blog/glassworm-returns-new-wave-openvsx-malware-expose-attacker-infrastructure): GlassWorm launches a new wave of attacks with updated tactics, our analysis reveals fresh attacker infrastructure now exposed.
- [Google and Microsoft Trusted Them. 2.3 Million Users Installed Them. They Were Malware.](https://www.koi.ai/blog/google-and-microsoft-trusted-them-2-3-million-users-installed-them-they-were-malware): Google and Microsoft trusted them, but 2.3 million users unknowingly installed malware, exposing data and systems to sophisticated attacks.
- [GreedyBear: 650 Attack Tools, One Coordinated Campaign](https://www.koi.ai/blog/greedybear-650-attack-tools-one-coordinated-campaign): GreedyBear’s campaign deployed 650 attack tools in a single coordinated operation, targeting global systems and exposing massive security risks.
- [Inside GhostPoster: How a PNG Icon Infected 50,000 Firefox Users](https://www.koi.ai/blog/inside-ghostposter-how-a-png-icon-infected-50-000-firefox-browser-users): Discover how GhostPoster used a malicious PNG icon to infect 50,000 Firefox users and the risks behind seemingly harmless downloads.
- [Koi Raises $48M to Reinvent Endpoint Security for the Modern Software Stack](https://www.koi.ai/blog/koi-raises-48m-to-reinvent-endpoint-security-for-the-modern-software-stack): Koi secures $48M to revolutionize endpoint security, delivering innovative protection for the modern software stack and enterprises.
- [How We Could’ve Taken Over Every Developer Using a VSCode Fork](https://www.koi.ai/blog/marketplace-takeover-how-we-couldve-taken-over-every-developer-using-a-vscode-fork-putting-millions-at-risk): Marketplace Takeover reveals how a VSCode fork could've compromised every developer, putting millions at risk of large-scale cyberattacks.
- [MCP Malware Wave Continues: A Remote Shell in Disguise](https://www.koi.ai/blog/mcp-malware-wave-continues-a-remote-shell-in-backdoor): A remote shell disguised as legitimate software powers the ongoing MCP malware wave, posing stealthy risks to countless systems.
- [Mining in Plain Sight: The VS Code Extension Cryptojacking Campaign](https://www.koi.ai/blog/mining-in-plain-sight-the-vs-code-extension-cryptojacking-campaign): Mining in Plain Sight exposes the VS Code extension cryptojacking campaign, silently hijacking developer machines for hidden cryptocurrency mining.
- [NPM Package With 56K Downloads Caught Stealing WhatsApp Messages](https://www.koi.ai/blog/npm-package-with-56k-downloads-malware-stealing-whatsapp-messages)
- [How 2.5 Billion Weekly Downloads Were Compromised in npm's Largest Supply Chain Attack](https://www.koi.ai/blog/npm-supply-chain-attack-package-poisoner): npm’s largest supply chain attack compromised 2.5 billion weekly downloads, exposing countless projects to risk and widespread vulnerabilities.
- [PhantomRaven: NPM Malware Hidden in Invisible Dependencies](https://www.koi.ai/blog/phantomraven-npm-malware-hidden-in-invisible-dependencies): PhantomRaven NPM malware hides in invisible dependencies, silently compromising projects and putting entire software supply chains at risk.
- [First Malicious MCP in the Wild: The Postmark Backdoor That's Stealing Your Emails](https://www.koi.ai/blog/postmark-mcp-npm-malicious-backdoor-email-theft): The first malicious MCP, named Postmark Backdoor, is stealing user emails by secretly intercepting and exfiltrating sensitive data.
- [PromptJacking: The Critical RCEs in Claude Desktop That Turn Questions Into Exploits](https://www.koi.ai/blog/promptjacking-the-critical-rce-in-claude-desktop-that-turn-questions-into-exploits): Critical RCE flaws in Claude Desktop (“PromptJacking”) let attackers turn simple user questions into dangerous exploits, risking data security.
- [Live Updates: Shai-Hulud, The Most Dangerous NPM Breach In History Affecting CrowdStrike and Hundreds of Popular Packages](https://www.koi.ai/blog/shai-hulud-npm-supply-chain-attack-crowdstrike-tinycolor): Live Updates: Shai-Hulud, The Most Dangerous NPM Breach In History Affecting CrowdStrike and Hundreds of Popular Packages
- [SpyVPN: The Google-Featured VPN That Secretly Captures Your Screen](https://www.koi.ai/blog/spyvpn-the-vpn-that-secretly-captures-your-screen): SpyVPN, once featured by Google, secretly captures users’ screens, compromising privacy and exposing sensitive data to surveillance risks.
- [The VS Code Malware That Captures Your Screen](https://www.koi.ai/blog/the-vs-code-malware-that-captures-your-screen): Malicious VS Code malware captures your screen activity, putting sensitive coding and personal information at serious risk for developers.
- [TigerJack's Extensions Continue to Rob Developers Blind Across Different Marketplaces](https://www.koi.ai/blog/tiger-jack-malicious-vscode-extensions-stealing-code): TigerJack’s malicious extensions are still stealing from developers across multiple marketplaces, putting sensitive projects and data in danger.
- [Trust Me, I’m Local: Chrome Extensions, MCP, and the Sandbox Escape](https://www.koi.ai/blog/trust-me-im-local-chrome-extensions-mcp-and-the-sandbox-escape): Trust Me, I’m Local reveals how Chrome extensions and MCP abuses enabled sandbox escapes, exposing users to major security vulnerabilities.
- [Trust Wallet Hack: Inside the Code That Stole $7M on Christmas Eve](https://www.koi.ai/blog/trust-wallet-binance-compromised-inside-the-code-that-stole-7m-on-christmas-eve): Technical analysis of the Trust Wallet Chrome extension attack reveals what the news got wrong - and how a simple policy would have prevented $7M in losses.
- [The NPM Malware That Tried to Gaslight Security Scanners](https://www.koi.ai/blog/two-years-17k-downloads-the-npm-malware-that-tried-to-gaslight-security-scanners): NPM malware with 17K downloads evaded scanners for two years, gaslighting security tools and risking developer supply chains.
- [8 Million Users' AI Conversations Sold for Profit by "Privacy" Extensions](https://www.koi.ai/blog/urban-vpn-browser-extension-ai-conversations-data-collection): Privacy browser extensions misled users and sold 8 million AI chat logs, exposing sensitive conversations for profit without consent.
- [VSCode Extension Trivia: Real or Cake?](https://www.koi.ai/blog/vscode-extension-trivia-real-or-cake): Is that VSCode extension trivia real or just a clever fake? Test your knowledge and spot the true tools from the imposters!
- [When Claude Becomes a Cyber-Weapon: The AI Arms Race Has Begun](https://www.koi.ai/blog/when-anthropic-claude-becomes-a-cyber-weapon-the-ai-arms-race-has-begun): AI like Claude can be repurposed as a cyber-weapon, fueling an escalating arms race in digital security and AI-driven attacks.
- [When Both Marketplaces Fall: The Cross-Platform Extension Malware Campaign](https://www.koi.ai/blog/when-both-marketplaces-fall-the-cross-platform-extension-malware-campaign): When both marketplaces fall, a cross-platform extension malware campaign exposes systemic trust failures across browsers globally!
- [When Chrome Extensions Turn Against Us: The Cyberhaven Breach and Beyond](https://www.koi.ai/blog/when-chrome-extensions-turn-against-us-the-cyberhaven-breach-and-beyond): Chrome extensions turned malicious in the Cyberhaven breach, exposing user data and highlighting new threats lurking in trusted add-ons.
- [WhiteCobra's Playbook Exposed](https://www.koi.ai/blog/whitecobra-vscode-cursor-extensions-malware): WhiteCobra’s critical mistake exposed its entire 24-extension campaign, targeting VS Code and Cursor users with stealthy persistent malware.
- [Live Updates: GlassWorm, First Self-Propagating Worm Using Invisible Code Hits OpenVSX And VSCode Marketplaces](https://www.koi.ai/incident/live-updates-glassworm-first-self-propagating-worm-using-invisible-code-hits-openvsx-and-vscode-marketplaces): Live Updates: GlassWorm, First Self-Propagating Worm Using Invisible Code Hits OpenVSX And VSCode Marketplaces
- [Live Updates: Sha1-Hulud, The Second Coming - Hundreds of NPM Packages Compromised](https://www.koi.ai/incident/live-updates-sha1-hulud-the-second-coming-hundred-npm-packages-compromised): A new wave of the Shai-Hulud malware is compromising hundreds of npm packages and destroying user home directories. Get live updates and mitigation steps.
- [Live Updates: Shai-Hulud, The Most Dangerous NPM Breach In History Affecting CrowdStrike and Hundreds of Popular Packages](https://www.koi.ai/incident/shai-hulud-npm-supply-chain-attack-crowdstrike-tinycolor): Live Updates: Shai-Hulud, The Most Dangerous NPM Breach In History Affecting CrowdStrike and Hundreds of Popular Packages