#Developer

I'm 21. Solo founder. Looking to connect with more developers & founders .

How Devin Covers Every Stage of the Software Lifecycle • Understand • Build • Test • Verify • Review • Secure • Operate Everything you need to know in 6 Minutes

Project is stable... as long as the tester is still holding it 😅🚛 #Developer #Tester

What if an AI video did not have to be a black box? SenseNova 6.8 Code2Video is an open-source model that generates video by writing code, giving developers editable source they can inspect, modify, and build on. It can analyze a reference video, recreate its visual structure and animation in code, and refine the result through multi-turn conversation without discarding the existing source. SenseTime is also opening the model weights, creation Skills, demo prompts, source code, and full creation traces. The video below is an official example created by SenseTime using SenseNova 6.8 Code2Video. GitHub:

We’re launching a preview of our Connected Apps Platform, giving developers a single place to build and manage apps for Search and the @GeminiApp. Plus, we’re rolling out a new set of Connected Apps from partners like @Adobe, @Calm, and @onepeloton to make your everyday activities more streamlined.

Congrats to @matsiiako, @dangtony98, @maidulll and @infisical (W23) on their $40M Series B! Infisical is the open source security infrastructure that manages API keys, passwords, and certificates for developers, machines, and AI agents. When an AI agent needs access to a sensitive system, Infisical hands it a fake credential and swaps in the real one through a proxy. The agent does its job without ever seeing the actual key. More than 1,300 organizations use the platform, including Airbus, Hugging Face, and the US Government, and they just passed 30,000 stars on GitHub.

The model weights are open. The economics of serving them are less so. OptimAI Router connects developers to independent Core Nodes. Choose a model through one API. Serve it from your computer. Set your price. Open AI should be open on both sides.

Introducing OptimAI Router 🌐 Open models have never been better. Using them is another story: running one yourself means hardware and upkeep, and renting one means someone else's prices and rules. Meanwhile, millions of capable computers sit idle every night. Router connects those two sides. 🛠 For builders • One OpenAI-style chat completions API, with streaming • You choose the model, and Router never swaps it • Try requests in the Playground before you write code • API keys with monthly caps, per-key limits and a max charge per request • Pay with OptimAI Credits and top up with x402 💻 For providers Set up in OptimAI Core Node in four steps: Computer → Model → Pricing → Serving. Earn credits for every completed request and collect Points while you serve. You set your price, and you start and stop serving whenever you like. Coming soon: MCP and x402 pay-as-you-go for individual requests. One network, two ways in. 👉

We built ChatGPT’s ad stack to respond to an evolving conversation, while ensuring answers stay independent. Here’s how the LLM orchestrator, retrieval, ranking and auction tools work together.

Top 10 Grok Bots for Developers Grab the one you need and try it 1. Cursor Projects Splits work into outer and inner loops. Inner loop is Cursor Projects. Outer loop is Grok Bot. Good template for new project bots. 2. Tripwire Tracks your repos, servers, domains, accounts and devices. Checks three times a day for real threats and only alerts when something affects you. 3. Security Guard Checks your PC once a week for known vulnerabilities in runtimes, certificates and tools. Only updates what you approve. 4. Ship Evidence Auditor Checks if AI-made releases actually meet the criteria using real evidence. Stops the release if something is wrong. 5. Gatekeeper Manages home Wi-Fi, router, devices and parental controls. Alerts you when unknown devices connect. 6. AEO/SEO Bot Pulls article ideas from YouTube, podcasts and calls. Drafts them and tracks rankings after publish. 7. Cloudflare Billing Bot Watches usage across your Cloudflare accounts and sends a daily summary so bills don’t spike unnoticed. 8. X Money Deep Report Bot Reads your X Money account weekly and sends a categorized report with cashback, interest and goals. 9. Video Creator Handles full explainer video production from research to upload using a director and specialist roles. 10. Starter Agent Coach Helps you build your first Grok Bot by asking one question at a time. Only drafts and always waits for your approval.

What do you do when a customer needs 1,000 Git repos created per second, and you have three weeks to deliver? You build your own Git server on Durable Objects. @dillon_mulroy breaks down how a small but mighty team of three shipped Cloudflare Artifacts as a closed beta, on time.

Want to see the HTML-in-Canvas API in action on a major production web app? 🖼️ @Shopify built their latest Shopify Editions using HTML-in-Canvas to bring interactive HTML components straight into canvas scenes. Visit the Editions site below and tap the "~" key on your keyboard to see the magic unfold before your eyes ⤵️

The new Edition drops today. It’s a magical time to build. Hope this Edition shows how much fun we’re having.

Harness Engineering from Scratch 🚀 Found this excellent 21K+ star GitHub repo that teaches you how to build reliable AI coding agents, from the ground up. An entire open-source curriculum with 14 lectures, 8 hands-on projects, and 8 progressive learning phases. What makes it useful is the approach: instead of just learning concepts, you gradually build the engineering environment around an AI agent, using the same real application throughout the course. What's inside: → Agent instructions and repository structure → Context management and persistent memory → Multi-session continuity and task handoffs → Task planning, scope control, and progress tracking → Testing, verification, and observability → Automated coding loops and maker-checker workflows → Graph orchestration, parallel execution, and rollback Includes practical projects, reusable templates, a harness-creator skill, and technical breakdowns of how Claude Code, Codex, Pi, and DeepSeek build their harnesses. Available in 15 languages. Perfect for software engineers and AI developers who want to move beyond prompting and understand how reliable coding agents are actually built. Completely free and open source. Here's the GitHub Repo:

Use your preferred coding agent directly in @AndroidStudio. You can now choose from over 40 coding agents inside the IDE, giving your agents project context and access to Android Studio tools 🧵👇

An Anthropic engineer reveals how he earns $350K–$950K: “I know how to build harnesses. 99% of developers still assemble multi-agent systems by hand, then struggle to keep them running reliably. I build systems that execute loops, operate inside sandboxes, and transform past logs into new skills. Same model. 10x more reliable agents that continuously improve, even while you sleep.” Opus 5.5 → Agents → Harnesses → Loops → Graphs In a short workshop, he explains how to build this system from the ground up: Step 1 → Control loops, context, and sandboxed tools with a harness Step 2 → Keep the model separate from its tools to generate the first token 60% faster Step 3 → Run agents on servers and recover from crashes using execution logs Step 4 → Handle failures with retries and replays instead of rebuilding everything from scratch Step 5 → Convert past execution logs into memories that help agents improve over time Anthropic calls this process “dreaming.” Most developers spend weeks piecing these systems together manually. This workshop shows you a more efficient approach. Bookmark this and watch it.

GOOD MORNING DEVELOPERS ⚡️✨ AI can understand words. But can it understand the world it is supposed to operate in? Imagine a robot entering a warehouse, navigating a crowded street, identifying a building entrance, or moving through an unfamiliar environment. Before it can perform these tasks effectively, it needs something essential: accurate information about the physical world. This is the problem @vangrid_io is working to solve. 🧠 WHAT IS @vangrid_io ? @vangrid_io is building a decentralized spatial data network designed to give Physical AI a better understanding of real-world environments. While traditional AI learns from digital information, robotics and autonomous systems also need detailed, ground-level representations of physical spaces. The challenge is that collecting this information can be expensive, time-consuming, and difficult to scale. Vangrid takes a different approach: turning everyday smartphones into tools for collecting spatial data. People capture real environments, those recordings are reconstructed into 3D models, and organizations can commission the specific locations and information they need. The bigger vision is to build a distributed spatial intelligence layer that connects human contributions with the growing demand for real-world data. 🌍 HOW DOES IT WORK? The system brings together three important participants. 1. Human contributors People use their smartphones to record physical environments, from buildings and warehouses to loading bays, walkways, and other locations. Instead of requiring every data collector to own specialized scanning equipment, Vangrid makes use of devices people already have. 2. Organizations that need spatial data Businesses and other requesters can publish funded bounties describing the locations they need captured and the rewards available. Contributors submit recordings, and the requester selects a suitable submission. Accepted captures can then be reconstructed into usable 3D models. This demand-driven approach helps connect data collection to specific requirements rather than generating activity without a clear purpose. 3. AI agents One of Vangrid’s most interesting ideas is enabling software agents to request physical-world data programmatically. An agent can identify a need, fund a capture request, and coordinate with a human contributor who can visit and record the environment. It creates a compelling connection between digital intelligence and physical action: AI identifies the need, humans capture the world, and the resulting data helps power intelligent systems. 🔐 WHY BLOCKCHAIN MATTERS Collecting data is only half the challenge. Organizations also need confidence in where that data came from and whether its recorded fingerprint has been altered. Vangrid uses cryptographic provenance to make capture records more transparent and verifiable. Its system groups capture fingerprints into Merkle trees and anchors the resulting records on Base through the Ethereum Attestation Service. This gives observers a way to verify that a capture fingerprint belongs to an attested record. The objective is to make spatial data more auditable rather than relying entirely on trust in the party providing it. Privacy is another important consideration. Vangrid documents an approach that blurs faces and vehicle registration plates on the recording device before footage is uploaded. 💰 TURNING REAL-WORLD DATA INTO AN ECONOMIC MARKET Vangrid connects spatial data collection with funded demand. Contributors receive payment when their submissions are accepted, subject to the bounty’s conditions and platform fees. The public bounty board currently displays more than $675,000 in settlements, providing a measurable indication of activity within its marketplace. @vangrid_io has also reported a $9 million seed funding round involving investors including HashKey Capital, Borderless Capital, Capital, and Animoca Brands.

Clear your calendar. Warm up your terminal. ☕ The Meta Global AI Developer Hackathon kicks off Oct 26: ten days to ship something awesome with our newest models, $150 in Meta Model API credits for every selected participant and a $1M prize pool. Applications close Oct 25. Limited to 20,000 selected applicants. git commit to it →

It is hard to believe, but there is now a way to avoid paying for Anthropic subscriptions. A project has appeared on GitHub that runs Claude Code completely for free It reroutes your traffic from Claude Code to 10 free providers, such as DeepSeek and Kimi, and over 21,000 developers are already using it Setup takes just 5 minutes This is definitely worth saving

The newly refreshed Devin Desktop app is now live! Full support for every Devin Cloud feature, a brand new agent command center, and thousands of improvements to how sessions are rendered. Download the new Devin Desktop app for free:

Today we're introducing a refreshed Devin Desktop app, with an agent command center designed to be smoother and more polished than ever before.