←
Codez1,956 posts
Codez avatar

Codez

@0xCodez

Content creator | AI researcher & builder | AI insights from 2030 | @beyond_xai

◷ Joined May 2026▧ 236 Photos and videos
Codez avatar
⟳ Codez retweeted
Codez@0xCodez·4h

These 20 GitHub repos will turn your Grok Bot into your most productive team ever This is f*cking gold. Skills. Team. Plan. Memory. Tools. 01 / Lauren Tan's noodle → Orchestrate agents using skills. github.com/poteto/noodle 02 / mattpocock/skills → Skills for real engineers, straight from his .agents directory. github.com/mattpocock/skills 03 / agent-skills → Production-grade engineering skills for AI coding agents. github.com/addyosmani/agent-… 04 / superpowers → An agentic skills framework and dev methodology that works. github.com/obra/superpowers 05 / Vercel agent-skills → Vercel's official collection of agent skills. github.com/vercel-labs/agent… 06 / awesome-agent-skills → 1000+ curated skills from official dev teams and the community. github.com/VoltAgent/awesome… 07 / Gas Town → Multi-agent workspace manager. github.com/gastownhall/gasto… 08 / gstack → 23 opinionated tools: CEO, designer, eng manager, release manager. github.com/garrytan/gstack 09 / BMAD Method → Agile, AI-driven development for agent teams. github.com/bmad-code-org/BMA… 10 / grokbot-field-notes → Rules, playbooks and bot roles from a 72-hour Grok Bot live build. github.com/unicodef1wn/grokb… 11 / Spec Kit → GitHub's toolkit for spec-driven development. github.com/github/spec-kit 12 / Task Master → AI-powered task management you can drop into your agent. github.com/eyaltoledano/clau… 13 / AGENTS.md → A simple, open format for guiding coding agents. github.com/agentsmd/agents.m… 14 / Beads → A memory upgrade for your coding agent. github.com/gastownhall/beads 15 / mem0 → The memory layer for AI agents. github.com/mem0ai/mem0 16 / Context7 → Up-to-date docs for AI agents. github.com/upstash/context7 17 / npx skills → The open agent skills tool. github.com/vercel-labs/skill… 18 / Composio → 1000+ toolkits with auth and tool search. github.com/ComposioHQ/compos… 19 / GitHub MCP Server → GitHub's official MCP server. github.com/github/github-mcp… 20 / Firecrawl → Search and scrape the web for your agents. github.com/firecrawl/firecra… Skill → plan → team → memory → ship. Bookmark this. Go build a team worth managing ↓

0:20
Codez@0xCodezAug 18

Codez @0xCodez Aug 18 Article Grok Bot Agents: how to automate your life in 10 Steps (Full-tutorial) Every AI tool you have used so far waits for you. You open it, you ask, it answers, you close it. The work still happens in your hands - the model just talks you through it. Follow my Substack to get

Higgsfield AI 🧩 avatar
⟳ Codez retweeted
Higgsfield AI 🧩@higgsfield·Nov 17

We built Higgsfield Miracle to address one of the biggest challenges in AI video: editing a specific part of a shot without changing the rest. Powered by GPT-6.2 Astra, it splits any video into layers for pixel-level editing in real time, with world knowledge built in. Every shot keeps the image depth that cinema-grade post-production needs. Now in private beta with select testers. Coming soon to Higgsfield MCP. Set up Higgsfield plugin in ChatGPT to be the first to try it.

1:32
Codez avatar
Codez@0xCodez·Nov 17

NVIDIA, AMD and Intel fight over every AI chip sold. Last week all three signed up for the same open source project. SGLang Summit, Nov 12, San Francisco: → SGLang Consortium: NVIDIA, AMD, Google, Meta, Intel, Thinking Machines Lab, Cloudflare, Modal and RadixArk join the maintainers to shape the roadmap → SGLang v1.0: built to run models in production at scale → SGLang now runs on AWS Trainium, co-developed with AWS Annapurna Labs → RadixArk x Intel: SGLang optimized for Intel CPUs, GPUs and accelerators, contributed back to open source → Miles v0.2: RL training on NVIDIA and AMD GPUs, early Google TPU support Why would rivals agree on anything? Because nobody wins if the software that runs the models belongs to one of them. The chip war is still on. Inference just became neutral ground.

Post attachment

2,000+ people. Two days. The first SGLang Summit. Together, we announced SGLang v1.0, released Miles v0.2, and launched the SGLang Consortium with nine founding members: @nvidia, @AMD, @intel, @Google, @Meta, @thinkymachines, @Cloudflare, @modal, and @radixark. SGLang is now used by 332 organizations. At Meta, it powers production inference for products serving billions of users. Teams from Modal, Cloudflare, LinkedIn, and others shared their production lessons. Across 90+ technical talks and six hands-on workshops, the community explored what comes next for inference, reinforcement learning, and AI agents. To everyone building, deploying, and contributing: thank you for making this possible. Explore the announcements, engineering lessons, and moments from SGLang Summit 2026 in the link below.

Post attachment
Higgsfield AI 🧩 avatar
⟳ Codez retweeted
Higgsfield AI 🧩@higgsfield_ai·Nov 17

GPT-6.2 Astra is LIVE on Higgsfield. OpenAI’s new frontier model for your most demanding projects. With Higgsfield plugin, GPT-6.2 Astra powers proactive multi-agent workflows across design, coding, and creative production. Join Higgsfield with up to 65% off for Black Friday.

1:15
OpenAI@OpenAINov 17

Some of you have been asking very politely. Some of you have just been asking. GPT-6.2 Astra and Sol are now available. Super efficient and built to perform where it counts—the real world. openai.com/index/introducing…

1:59
Codez avatar
⟳ Codez retweeted
Codez@0xCodez·Nov 17

GrokBot team just dropped 1-hour workshop on running your work with AI agents from 0% to 100%: “GrokBot is not a chatbot” • 10% → 0:57 - how GrokBot team manage usage limits • 30% → 6:41 - how to build skills for GrokBot • 55% → 10:41 - building Chief of staff agent • 80% → 12:48 - GrokBot orchestrator for Cursor agents • 100% → 27:01 - how to replace your software with bots most people still ask it questions like a chatbot - the team runs it as a teammate that directs other agents watch it today - then read the full GrokBot playbook below ↓

1:17:51
Codez@0xCodezAug 18

Codez @0xCodez Aug 18 Article Grok Bot Agents: how to automate your life in 10 Steps (Full-tutorial) Every AI tool you have used so far waits for you. You open it, you ask, it answers, you close it. The work still happens in your hands - the model just talks you through it. Follow my Substack to get

Higgsfield AI 🧩 avatar
⟳ Codez retweeted
Higgsfield AI 🧩@higgsfield·Nov 16

10 days left to enter the $1,000,000 Higgsfield Talent Contest. How to enter: > Make a 3-7 min film, any genre or style > Submit by Nov 26 Build it in public on Higgsfield → standout projects earn up to 100K credits. Now with up to 65% off for Black Friday.

1:08
Codez avatar
⟳ Codez retweeted
Codez@0xCodez·Nov 16

SpaceXAI engineer, Lauren Tan: "All code is slop, except mine.i’m running a team of 20+ GrokBot agents, shipping code even while I sleep agents verify their own work with videos, screenshots and tests" in 17-minutes at AI Code Summit she just showed her entire coding setup. Every agent mistake becomes a lint rule, not a prompt. Result: 3,000 PRs to production in a month. Her codebase is now clean enough that agents merge most of their own PRs. Watch this before you automate anything. Then read full setup in article below.

17:18
Codez@0xCodezAug 18

Codez @0xCodez Aug 18 Article Grok Bot Agents: how to automate your life in 10 Steps (Full-tutorial) Every AI tool you have used so far waits for you. You open it, you ask, it answers, you close it. The work still happens in your hands - the model just talks you through it. Follow my Substack to get

Codez avatar
⟳ Codez retweeted
Codez@0xCodez·Nov 15

SpaceXAI engineer, Lauren Tan just dropped a free 1-hour masterclass on build team of GrokBot agents and loops from 0% to 100%: “I’ve gotten to a point where i essentially have my agents merge their own pull requests” 10% → 3:36 - giving bots real jobs: own logins, budget 30% → 22:35 - the workflow behind 4,000 PRs a month 55% → 29:01 - building a personal software factory 80% → 40:13 - the smallest loop: teach agents to check their own work 100% → 47:33 - codebase guardrails that save tokens most people are still babysitting one agent - Lauren runs a factory that ships 4,000 PRs a month and fixes bugs at 3 AM bookmark & watch - then read the full playbook on GrokBot team below ↓

54:56
Codez@0xCodezAug 18

Codez @0xCodez Aug 18 Article Grok Bot Agents: how to automate your life in 10 Steps (Full-tutorial) Every AI tool you have used so far waits for you. You open it, you ask, it answers, you close it. The work still happens in your hands - the model just talks you through it. Follow my Substack to get

Codez avatar
Codez@0xCodez·Nov 15

Can an AI rediscover a planet? 155 years of Uranus data. No internet. A secretly rotated sky. 2.5 hours and ~300 steps of its own orbital-mechanics code later, IQuest-Q1 told me where to look. Neptune was 9.6° away. 🔭 What's under the hood: → Sparse Mixture-of-Experts: big model, light inference → Native long context, so a 2.5-hour session stays in one run → Recursive Drafter: speculative decoding that cuts long-sequence cost → Built from scratch by IQuest Research, open weights huggingface.co/IQuestLab/IQu…

1:56
Codez avatar
⟳ Codez retweeted
Codez@0xCodez·Nov 14

SpaceX paid $60B for Cursor. the next $1T company won't hire 100 people. It will run 24 agents. this is the 6-layer stack I'd run on GrokBot + SpaceXAI compute. You sell the outcome at an agency price. You run it at the cost of a SaaS. [1] Research → Grok DeepSearch mines X for real pain points → Map the market + tear down every competitor → Plan Mode turns the findings into a build spec [2] Build → One agent turns the spec into code → Background agents ship tasks in parallel → Composer handles the multi-file edits [3] Ship → Bugbot reviews every single PR → Tests + preview deploys run on their own → You approve. Agents merge to prod. [4] Grow → Grok drafts threads in your voice → Grok Imagine cuts the launch videos → A GrokBot replies to every engager [5] Revenue → Onboarding runs as one workflow → Support bot closes tickets 24/7 → Chief of Staff bot sends you a daily brief [6] Brain + compute → Every PR, doc, call and ticket gets indexed → Linked into 1 knowledge graph every agent reads → Runs on Colossus-scale GPU compute Your job shrinks to 2 things. Pick the problem. Approve the PR. Most founders will keep hiring a person for every layer. The ones who wire agents into it will run the whole company solo. Save this post for later. Copy-paste prompts for every layer in article below ↓

0:16
Codez@0xCodezAug 18

Codez @0xCodez Aug 18 Article Grok Bot Agents: how to automate your life in 10 Steps (Full-tutorial) Every AI tool you have used so far waits for you. You open it, you ask, it answers, you close it. The work still happens in your hands - the model just talks you through it. Follow my Substack to get

Codez avatar
⟳ Codez retweeted
Codez@0xCodez·Nov 13

Google Brain founder, Andrew Ng: "Building swarms of 100+ agents with self-improving loops & graphs is a must skill for future AI engineers 90% of my work is already automated by agentic loops" his team went through 20,000+ job descriptions to find 5 skills AI engineers actually get paid for: 1/ directing agents > writing code. Ng’s own pandas and JS syntax is fading, and he doesn’t miss it. Simple apps: let the agent choose. Demanding ones: whoever steers the trade-offs wins. 2/ make your data agent-ready. Half of any AI project is still data cleanup and plumbing. Mature that agent-ready layer once, and every project after it runs twice as fast. 3/ building on uncertain outputs. Old software engineering optimized for correct code. AI apps break that - output is non-deterministic by default. The skill is shipping something reliable on top of something that isn’t. 4/ fundamentals are how you steer. Without them, the agent picks a random point on the cost / latency / reliability / security spectrum. If you don’t know a trade-off exists, you can’t tell the agent to make it right. 5/ shaping the build, not the spec. The rarest skill is deciding what gets built. humans carry vast context agents don’t have - the customer’s face when they hate a feature. That advantage persists. this 20-minute talk is the best breakdown of what’s coming in AI engineering. watch it today, then read the article below on how to build and ship agentic graphs and loops from scratch.

22:00
Codez@0xCodezJul 20

Codez @0xCodez Jul 20 Article Graph Engineering with Claude: 14-Step roadmap from 0 to graph architect (Full Course) Most people who try to build a multi-step agent end up with a straight line. Step one, step two, step three - each waiting politely for the last to finish before it starts. 9/10 notice that half those

Codez avatar
Codez@0xCodez·Nov 12

i used fable 5.2 to make this "hand-painted" wizard game in one evening this f*cking insane... every frame is about 1.5 million brush strokes, running in real time in a browser tab. Merlin strikes his staff and the whole world repaints itself: spring → autumn → winter → a glowing dream.

0:43
Movez@0xMovezNov 12

Movez @0xMovez Nov 12 Article How to build a $1M AI animation studio with Fable 5.2 + Picsart Most people who try to make an AI series end up with one great clip and a stranger with the same name in the next one. They render before they direct. They pay for 1080p before they've seen a draft.

Codez avatar
Codez@0xCodez·Nov 12

ex. Jane Street Quant ~$400k/year just revealed Fable 5.2 & Opus 5.5 trading system this shouldn’t be f*cking public... It turns Claude into 24/7 research & trading swarm of AI agents Send it to Opus 5.5 & Fable 5.2, and enjoy. then read how to build your personal AI hedge fund with @horizon_trade_x below ↓

Post attachment

Introducing Horizon Your next investment starts with a thesis Nvidia just lost its CEO. Bullish or bearish? Tell Horizon It builds your strategy and runs it live in your broker Or copy strategies from top investors with real track records Have an audience? Publish and get paid

0:44