🐝 OpenHive v0.2.21 is live. Three things shipped this version: 🌥️ Open Cloud Computer: one-click handoff from any colony to a cloud runtime. No more "my laptop has to stay on." 📜 Paged chat loading: colony sessions with thousands of messages now stay smooth on first paint. 💼 LinkedIn Premium skills: Sales Navigator-aware sourcing, ICP scoring, InMail handling, and a new scan post reactors skill that powers the engagement → personalized-DM loop. Plus the boring-but-important stuff: hardened browser bridge lifecycle, smoother scrolling, Windows install reliability, and credits now have a daily allowance on top of monthly. Check out the full changelog: https://lnkd.in/gaZxgR5Y
OpenHive
Technology, Information and Internet
San Francisco, California 5,419 followers
Deploy multi-agents swarm to run complex, long-session business processes.
About us
Aden is the creator of OpenHive, the first vertically integrated cloud operating system and hypervisor engineered natively for autonomous AI agents. While legacy clouds were built for stateless, human-operated microservices, OpenHive provides the persistent memory, semantic state recovery, and secure isolation required by continuous digital labor.
- Website
-
https://open-hive.com/
External link for OpenHive
- Industry
- Technology, Information and Internet
- Company size
- 51-200 employees
- Headquarters
- San Francisco, California
- Type
- Privately Held
- Founded
- 2025
- Specialties
- Infrastructure, Agentic AI, Virtualization, Hypervisor, Virtual Desktop Infra (VDI), Distributed System, Baremetal Provisioning, Enterprise IT Security, Sandboxing & Execution Isolation, and Infrastructure as a Service (IaaS)
Locations
-
Primary
Get directions
San Francisco, California 94107, US
Employees at OpenHive
Updates
-
OpenHive reposted this
𝗜 𝗮𝘁𝘁𝗲𝗻𝗱𝗲𝗱 𝗮 𝘀𝗲𝘀𝘀𝗶𝗼𝗻 𝗲𝘅𝗽𝗲𝗰𝘁𝗶𝗻𝗴 𝘁𝗼 𝘁𝗮𝗸𝗲 𝗻𝗼𝘁𝗲𝘀. 𝗜𝗻𝘀𝘁𝗲𝗮𝗱 𝗜 𝗲𝗻𝗱𝗲𝗱 𝘂𝗽 𝘀𝗵𝗶𝗽𝗽𝗶𝗻𝗴 𝗰𝗼𝗱𝗲 𝘁𝗼 Amazon Web Services (AWS). A few weeks ago I joined an AgentCore Observability session hosted by Madhu Samhitha Vangara, Pete Ildefonso IV, and Rajesh Kumar Ravi. Something clicked. I wanted to 𝘂𝗻𝗱𝗲𝗿𝘀𝘁𝗮𝗻𝗱 𝗵𝗼𝘄 𝗮𝗴𝗲𝗻𝘁𝗶𝗰 𝗔𝗜 systems actually work under the hood, not just use them. 𝗦𝗼 𝗜 𝘀𝘁𝗮𝗿𝘁𝗲𝗱 𝗰𝗼𝗻𝘁𝗿𝗶𝗯𝘂𝘁𝗶𝗻𝗴: First, Shinzo Labs, where I worked on AI agent MCP observability, the exact topic that session got me curious about. Then LiteLLM (YC W23), where I contributed to the gateway that routes 100+ LLM APIs including AWS Bedrock. Then Aden AI, digging into agent framework internals. 𝗘𝗮𝗰𝗵 𝗿𝗲𝗽𝗼 𝘁𝗮𝘂𝗴𝗵𝘁 𝗺𝗲 something different about how agents are built, observed, and scaled. Today I shipped 𝟯 𝗣𝗥𝘀 𝘁𝗼 𝗮𝘄𝘀𝗹𝗮𝗯𝘀/𝗮𝗴𝗲𝗻𝘁-𝘀𝗾𝘂𝗮𝗱, AWS's own open-source multi-agent orchestrator: 1. Fixed a silent exception bug in ChainAgent that was swallowing errors. 2. Built a full GeminiAgent implementation bringing Google's models into the AWS multi-agent ecosystem. 3. Fixed a streaming response bug in SupervisorAgent that broke multi-turn conversations. 𝗢𝗻𝗲 𝘀𝗲𝘀𝘀𝗶𝗼𝗻 turned into 𝗳𝗼𝘂𝗿 𝗼𝗽𝗲𝗻-𝘀𝗼𝘂𝗿𝗰𝗲 𝗽𝗿𝗼𝗷𝗲𝗰𝘁𝘀. The agentic AI ecosystem is moving fast and the 𝗯𝗲𝘀𝘁 𝘄𝗮𝘆 𝘁𝗼 𝗹𝗲𝗮𝗿𝗻 𝗶𝘁 𝗶𝘀 𝘁𝗼 𝗯𝘂𝗶𝗹𝗱 𝗶𝘁. https://lnkd.in/gmCazCcm https://lnkd.in/gvjde5JU #AWS #OpenSource #AI #GenAI #AgenticAI #Bedrock #LLM #MultiAgentSystems #AIAgents
-
🐝 Hive v0.6.6 is here - 60+ Integrations 🐝 big one today - queen got a brain upgrade 👑 Planning phase Queen now thinks before she builds. She'll explore your codebase, discover tools, design the architecture with you, and only start coding after you say go. No more yolo builds 🧠 Memory She remembers things now. across sessions. keeps a diary in ~/.hive/queen/ - what you built together, what worked, what didn't. Next time you open hive she picks up where you left off full notes → https://lnkd.in/gpkTpyig Thank you to everyone who contributed to this release @akshajtiwari @Amdev-5 @Waryjustice @karthik-kotra
-
-
We are excited to have you Esther Tsotso
I’m excited to share that few weeks ago, I received my first opportunity to contribute to open source. Right in the middle of thinking about where to even begin with OSC, I received a message from Vincent Jiang through Y Combinator inviting me to contribute to Aden Aden is building a development platform for self-evolving AI agents capable of running business processes autonomously, which is a fascinating space to explore as AI continues to reshape how software operates. I’ll be contributing part-time alongside school and my personal projects, working with a broader community of open-source enthusiasts and builders from around the world. This will also be my debut into open-source contributions, so there’s a mix of excitement and a bit of nervousness — but mostly curiosity about what I’ll learn and who I’ll get to collaborate with. I’m really looking forward to what this experience brings. If you’ve contributed to open source before, I’d love to hear your advice for getting started. #OpenSource #SoftwareEngineering #AI #BuildInPublic #Developers
-
🚀 Aden Hive v0.6.2 is out This release is all about making agent workflows clearer, more modular, and more production-ready: What’s new in v0.6.2 🐝 Queen & Worker inputs are fully separated (no more routing confusion) 🐝 New Subagent framework with lifecycle tools + progressive feedback 🐝 GCU (General Computer Use) is now a first-class node type → built-in browser automation 🔥 Pull latest main and try it: https://lnkd.in/ezcT7NpR If you’re building agents in real workflows, I’d love you to check it out. Shoutouts to our top contributors: https://lnkd.in/gchcQ3RC
-
-
Love this. Appreciate you taking a serious infra lens here. You nailed the real gap: demos are easy, production is chaos: reliability/observability/state are the whole game. Excited to see your thoughts on determinism and long-running persistence as you dig in Muhammad Ukasha
🔥Exploring Agent Infrastructure at a Y Combinator-Backed AI Startup🔥 Recently connected with Vincent Jiang and the team at Aden (Y Combinator backed startup) and was invited to contribute to their open-source project, Hive. 👉Hive is a development platform for generating self-evolving AI agents that execute business processes autonomously. One thing that consistently stands out: AI agents perform well in controlled demos — but production environments introduce real-world complexity. The gap between experimentation and enterprise systems often comes down to: • Reliability • Observability • State management 👉As I begin exploring the architecture, I’m particularly interested in workflow determinism, long-running task persistence, and cost-aware orchestration. Looking forward to contributing and sharing insights as I dive deeper into autonomous infrastructure. Check out the project here: https://lnkd.in/d2kBDBWZ #OpenSource #AIAgents #Automation #YCombinator #AIStartup #BuildingInPublic #AdenHive
-
-
So happy to have you as a contributor Amit Kumar
📸 Big News for Our Hive agent coder 📸 In later today, we’re dropping the latest release - the Hive Coding Agent is going live. - No more Claude-code dependency. - No more model lock-in. - No more workarounds. We now have a coding agent that understands Hive 100% - and you can power it with any LLM model you prefer. Your workflow, your choice. It will be a Hive coder agent building Hive worker agent for you. And to celebrate, we’re giving FREE LLM key to our top contributors. This is our way of saying thank you to the builders who’ve been pushing Hive forward every day. We see you. We appreciate you. And we want to make it even easier for you to build with us. Shoutouts to our top contributors: https://lnkd.in/gchcQ3RC
-
-
Happy to have you on our team, Satvik Garg!
I am excited to share that I am now an open-source contributor at Aden (via YC Startups), where I am working directly on their core execution engine. Aden is developing an outcome-driven execution engine for autonomous AI agents. This innovative framework moves away from brittle, static Directed Acyclic Graphs (DAGs) and instead emphasizes self-healing runtimes and adaptive logic. Recent contributions to the core framework include: - Clarified evaluation semantics in the graph execution layer (OutcomeAggregator), making success criteria explicit. - Built new tool integrations (MCP) with full credential handling and unit tests. - Navigated strict CI pipelines (make check, pytest) and rigorous architectural review cycles. - Gained solid hands-on exposure to reliability engineering, explicit state management, and production-grade infrastructure. For more details, you can check out : https://adenhq.com/
-
📸 Big News for Our Hive agent coder 📸 In later today, we’re dropping the latest release - the Hive Coding Agent is going live. - No more Claude-code dependency. - No more model lock-in. - No more workarounds. We now have a coding agent that understands Hive 100% - and you can power it with any LLM model you prefer. Your workflow, your choice. It will be a Hive coder agent building Hive worker agent for you. And to celebrate, we’re giving FREE LLM key to our top contributors. This is our way of saying thank you to the builders who’ve been pushing Hive forward every day. We see you. We appreciate you. And we want to make it even easier for you to build with us. Shoutouts to our top contributors: https://lnkd.in/gchcQ3RC
-