โ†
Every ๐Ÿ“ง@every
open on x โ†—

Inside Every's AI engineering stack: @kieranklaassen plans every @CoraComputer feature at three levels before touching code himselfโ€”small (one-shot), medium (needs review), large (manual work required). His edge: Context 7 MCP, a tool that pulls up-to-date, version-specific docs and code examples from official sources directly into AI prompts. Result: agents stay grounded in reality instead of hallucinating outdated solutions. Full breakdown: every.to/source-code/inside-the-ai-workflows-of-

โ™ฅ 47โ‡„ 4โ—‰ 6.9K
#claude-code
#coding-agents
pinned 20 OCT 24ยท backfill

related

I took the @karpathy autoresearch loop and pointed it at markets. 25 AI agents debate macro, rates, commodities, sectors, and single stocks daily. Every recommendation scored against real outcomes. Worst agent by rolling Sharpe gets its prompt rewritten by the system. Keep or revert. Same loop, prompts are the weights, Sharpe is the loss function. Trained the agents on 18 months of market data. 378 iterations. 54 prompt modifications, 16 survived. The system learned which agents to trust using Darwinian weights โ€” geopolitical, commodities, and the @BillAckman quality compounder rose to the top. The agents even figured out their own portfolio manager was the weakest link before we did! Deployed the trained agents. +22% in 173 days. Best pick: AVGO at $152, held for +128%. The final prompts are evolutionary products โ€” shaped by market feedback, not human intuition. Now running live with my own capital. https://github.com/chrisworsey55/atlas-gic Part hedge fund, part research experiment :)@Chris_Worsey@flowisgreat_@tom_doerrโ–ถ@coderabbitaiIntroducing eve, an agent framework. ๐šŠ๐š๐šŽ๐š—๐š/ ๐šŠ๐š๐šŽ๐š—๐š.๐š๐šœ ๐š’๐š—๐šœ๐š๐š›๐šž๐šŒ๐š๐š’๐š˜๐š—๐šœ.๐š–๐š ๐š๐š˜๐š˜๐š•๐šœ/ ๐šœ๐š”๐š’๐š•๐š•๐šœ/ ๐šœ๐šŠ๐š—๐š๐š‹๐š˜๐šก/ ๐šœ๐šŒ๐š‘๐šŽ๐š๐šž๐š•๐šŽ๐šœ/ Like Next.js, for agents. https://vercel.com/blog/introducing-eve@vercel@Zai_org@interjc@dhruvtwt_โ–ถ@_chenglou
@chrstnerodeโ‰ก 11+11@haydenbleaselโ–ถ@ctatedevVercel Ship (so far) โ–ฒ Vercel Services โ€” Run full stack on Vercel โ€” Microservices on Vercel โ€” Private service comms โ–ฒ eve โ€” The framework for building agents โ€” Durable agents with one folder โ€” Sandboxing, approvals and evals โ–ฒ WebSocket support โ€” Persistent connections โ–ฒ Vercel Functions โ€” Dockerfile support โ€” 5 GB packages, 30 min โ€” Zero-config Node servers โ–ฒ Vercel Agent in Dashboard โ€” Autonomous investigations โ€” Plan-based permissions โ€” Fixes surfaced as PRs โ–ฒ Vercel Container Registry โ€” Container images on Vercel โ€” OCI push, pull and tag โ–ฒ Vercel Sandbox โ€” Custom images โ€” Persistent sandbox storage โ€” 24 hour runtime + expiration โ–ฒ AI SDK 7 + agent harnesses โ€” Toolkit for building agents โ€” Coding-agent harnesses โ–ฒ Chat SDK โ€” Multi-channel agents โ–ฒ Access & security โ€” Vercel Passport: OIDC access โ€” Vercel Connect: external services โ€” OIDC audiences + Audit Logs โ–ฒ Enterprise apps and agents โ€” Security posture dashboard โ€” Vercel Functions in your cloud โ–ฒ AI Gateway โ€” Voice, speech and transcription โ€” API key budgets โ–ฒ Vercel Blob โ€” Private blob storage โ€” Signed URLs and OIDC auth โ–ฒ Builds & deployments โ€” 500 concurrent builds for Pro โ€” Vercel Drop โ€” Elastic builds โ–ฒ Vercel CLI โ€” Speed Insights + Web Analytics โ€” Trace any request โ€” Deployment limits removed โ–ฒ Observability โ€” Trace eve sessions โ€” Trace Workflow runs โ–ฒ Workflows โ€” Inflight cancellation โ€” Nitro v3 + TanStack Start โ–ฒ SDK-free Vercel Flags โ€” No SDK keys@ctatedevโ–ถ@CopilotKit@emilkowalski@techwith_ramLow-key websites I quietly rely on 1) http://roadmap.sh Gives you a brutally clear learning path for roles like frontend, backend, DevOps, etc No fluff, just โ€œlearn this โ†’ then this โ†’ then thisโ€. 2) http://playcode.io An online playground to quickly test HTML, CSS, JS without setting up anything locally Perfect for quick experiments and debugging ideas 3) http://usehooks.com A collection of reusable React hooks with real use cases Saves time and helps you avoid rewriting the same logic again and again 4) http://devhints.io Concise cheat sheets for languages, frameworks, and tools. Ideal when you forget syntax and donโ€™t want to read a 20-minute blog 5) http://jsoncrack.com Turns messy JSON into a clean visual tree Makes understanding large APIs and configs way easier than staring at raw text 6) http://realtimecolors.com Lets you generate and preview color palettes instantly Useful when you want decent UI colors without guessing or copying blindly 7) http://regex101.com Build, test, and debug regex step by step with explanations Honestly, the fastest way to stop hating regex 8) http://bundlephobia.com Shows how big an npm package really is before you install it Helps you avoid bloating your app with โ€œtinyโ€ libraries 9) http://caniuse.com Tells you which CSS/JS features actually work across browsers Essential before using shiny new features in production 10) http://toolbox.googleapps.com Googleโ€™s own diagnostics tools for DNS, email, headers, and network issues Surprisingly useful for debugging real-world problems ๐Ÿ‘‰ Which one of these do you already use and which one did you not know existed?@shekhu04@PrajwalTomar_@dr_cintasAnnouncing Paper Desktop + MCP Now any agent can read or write with Paper Paper is connected. To your code. To your agents. To your data. To your teammates. Download Paper Desktop at http://paper.design@paper