AI agents often fail because their instructions, or skills,... are manually modified with no guarantee of improvement. Learn how SkillOpt turns skill editing into a training process, making agent behavior more reliable without changing model weights:show more

Important new course: Agent Skills with Anthropic, built with... Anthropic and taught by Elie Schoppik! Skills are constructed as folders of instructions that equip agents with on-demand knowledge and workflows. This short course teaches you how to create them following best practices. Because skills follow an open standard format, you can build them once and deploy across any skills-compatible agent, like Claude Code. What you'll learn: - Create custom skills for code generation and review, data analysis, and research - Build complex workflows using Anthropic's pre-built skills (Excel, PowerPoint, skill creation) and custom skills - Combine skills with MCP and subagents to create agentic systems with specialized knowledge - Deploy the same skills across Claude Code, the Claude API, and the Claude Agent SDK Join and learn to equip agents with the specialized knowledge they need for reliable, repeatable workflows.show more

Coding agents that actually get better the more your... team uses them. Introducing Hivemind: continual learning for AI coding agents. Hivemind turns the traces from every agent your team runs (Claude Code, Codex, Cursor, Hermes, OpenClaw, Pi) into reusable skills, then pushes those skills across all of them. All on your cloud storage. Now with SkillOpt built in, your skills get trained: +19.1 points of accuracy in Claude Code, +24.8 in Codex, best or tied on all 52 setups tested. Open source, one line install.show more

You need to try Hermes RIGHT NOW. OpenClaw was... the breakthrough. Hermes feels like the evolution. Here’s why AI power users are quietly switching: • Hermes learns from you over time • It creates its own reusable skills • Memory stays curated instead of bloated • Agents improve every 10 turns automatically • It actually feels stable on day 30 The wildest part: Hermes agents build workflows from your behavior. One user asked their agent to configure Twingate once. The agent: → learned the process → created a reusable skill → stored it for future tasks → improved its own workflow automatically That’s not prompt engineering anymore. That’s agent evolution. Meanwhile most people are still babysitting broken OpenClaw setups and manually importing marketplace skills. Hermes took a different path: Less clutter. Less tweaking. More autonomy. And it’s working. It already flipped OpenClaw in OpenRouter token usage and became one of the fastest-growing AI repos on GitHub. The bigger shift: We’re entering the era where the model matters less than the harness around it. GPT-5.5, Grok, Qwen: They’re already powerful enough. The real edge now is: • memory systems • self-improvement loops • skill creation • agent orchestration Hermes understands that. Most people still think AI agents are glorified chatbots. The people using Hermes are building AI teammates.show more

Today, we’re shipping new ways to observe, analyze, and... debug agents with LangSmith: • Polly: an AI assistant for AI engineering that helps you understand traces, threads, and improve prompts • LangSmith Fetch: a CLI for pulling trace & thread data straight into your terminal or coding agent Agents are running longer and getting more complex, which demands new debugging workflows beyond simple LLM apps. We wrote a blog on the trends behind this shift— and why tools like Polly and LangSmith Fetch are needed. Shipping reliable agents requires full visibility into agent behavior, with tooling that helps you reason over that data in the UI, the terminal, or alongside coding agents. 📔Learn more about Polly: 📔Learn more about LangSmith Fetch: 📔How to observe deep agents: 📽️ Polly video tutorial: 📽️ LangSmith Fetch video tutorial:show more

New course: Nvidia's NeMo Agent Toolkit: Making Agents Reliable,... taught by Brian McBrayer 🐬 from NVIDIA. Many teams struggle to turn agent demos into reliable systems that are ready for production. This short course teaches you to harden agentic workflows into reliable systems using Nvidia's open-source NeMo Agent Toolkit (NAT). Whether you built your agent in raw Python or using a framework like LangGraph, or CrewAI, NAT provides building blocks for observability, evaluation, and deployment that turn proofs-of-concept into production-ready systems. NAT makes it easy to troubleshoot and optimize agent performance with execution traces, systematic evaluations, and CI/CD integration. Skills you'll gain: - Build configuration-driven agent workflows with REST APIs and minimal code - Add observability with tracing to visualize agent reasoning and debug performance bottlenecks - Create systematic evaluations using gold-standard datasets to measure and improve agent reliability - Deploy multi-agent systems with authentication, rate limiting, and professional web interfaces - Orchestrate agents from different frameworks to collaborate on complex tasks Join and learn how to turn agent demos into reliable systems!show more

Most AI agents do not fail because the prompt... is weak. They fail because there is no loop around the prompt. A real loop: - Finds work - Executes - Verifies - Saves state - Stops or escalates If an agent can say “done” without tests, a budget cap and an independent verifier, you built a demo. Prompts make you the operator. Loops make the agent useful while you are offline.show more

New Short Course: Building AI Browser Agents! Learn how... to build AI agents that interact and take actions on websites in this course, created in partnership with and taught by and @namangarg0, Co-founders of AGI Inc. AI browser agents can log into websites, fill out forms, click through web pages, or even place orders online for you. They use both visual information, like screenshots, and structural data, like the HTML or Document Object Model (DOM) of a web page, to reason and take action. With the complexity of webpages and multiple possible actions at each step, it can be challenging for an AI browser agent to complete an assigned task. Because these agents run long action sequences, a single error—like clicking the wrong button or misreading a field—can lead to unexpected outcomes or errors that compound over time. In this course, you'll understand how autonomous web agents work, their current limitations, and how AgentQ enables them to improve through self-correction. In detail, you'll: - Learn what web agents are, how they automate tasks online, their architecture, key components, limitations, and an overview of their decision-making strategies. - Build a web agent that can scrape website and return course recommendations in a structured output format. - Build an autonomous web agent that can execute multiple tasks, such as finding and summarizing webpages, filling out a form, and signing up for a newsletter. - Explore AgentQ, a framework that enables agents to self-correct by combining Monte Carlo Tree Search (MCTS), a self-critique mechanism for continuous improvement, and Direct Preference Optimization (DPO). - Deep dive into MCTS, learn how it finds an effective path, illustrated by an example of Gridworld animation, and use AgentQ to complete web tasks. - Understand AI agents' current state and future directions—including key factors shaping their evolution, such as hardware, algorithm innovation, and data availability. By the end of this course, you will have hands-on experience building browser agents and a deeper understanding of how to make them more robust and reliable. Please sign up here:show more

If a model is 50% reliable, it's 0% useful.... Michael Giannangeli, Head of Agentic AI at Amazon Nova, breaks down one of the hardest problems in agentic AI: reliability. Models need to match or beat human performance, or they won't get adopted. To close the gap, teams are building an ecosystem of reinforcement learning "gyms" where AI agents can hone their skills. Think of it like a virtual training ground where agents practice tasks through trial and error, building reliability one rep at a time. 💪show more

AI video creation is entering the Skill era. In... the past, making a video like this meant complex prompting, constant trial and error, and moving between separate tools for visuals, video, music, editing, and refinements. Now, with Skills, the creative workflow can be reused. The pacing, camera direction, scene structure, visual style, and generation process are no longer one-off instructions. They become part of the system.show more

today, we are making the Mosaic video editing api... available to all agents & humans. see how we setup larry — our slack openclaw agent — with a mosaic api key and had it clip, edit, and post archived steve jobs footage. all without ever leaving slack. you can connect your openclaw agents, claude code, or vibe-coded saas to our agentic video editing api. no waitlist — now live at mosaic [dot] so. comment "API" to get 100 free credits dropped into your account to get started with the api for free. more details on why & how we’re making this change below (thread):show more

OpenClaw meets RL! OpenClaw Agents adapt through memory files... and skills, but the base model weights never actually change. OpenClaw-RL solves this! It wraps a self-hosted model as an OpenAI-compatible API, intercepts live conversations from OpenClaw, and trains the policy in the background using RL. The architecture is fully async. This means serving, reward scoring, and training all run in parallel. Once done, weights get hot-swapped after every batch while the agent keeps responding. Currently, it has two training modes: - Binary RL (GRPO): A process reward model scores each turn as good, bad, or neutral. That scalar reward drives policy updates via a PPO-style clipped objective. - On-Policy Distillation: When concrete corrections come in like "you should have checked that file first," it uses that feedback as a richer, directional training signal at the token level. When to use OpenClaw-RL? To be fair, a lot of agent behavior can already be improved through better memory and skill design. OpenClaw's existing skill ecosystem and community-built self-improvement skills handle a wide range of use cases without touching model weights at all. If the agent keeps forgetting preferences, that's a memory problem. And if it doesn't know how to handle a specific workflow, that's a skill problem. Both are solvable at the prompt and context layer. Where RL becomes interesting is when the failure pattern lives deeper in the model's reasoning itself. Things like consistently poor tool selection order, weak multi-step planning, or failing to interpret ambiguous instructions the way a specific user intends. Research on agentic RL (like ARTIST and Agent-R1) has shown that these behavioral patterns hit a ceiling with prompt-based approaches alone, especially in complex multi-turn tasks where the model needs to recover from tool failures or adapt its strategy mid-execution. That's the layer OpenClaw-RL targets, and it's a meaningful distinction from what OpenClaw offers. I have shared the repo in the replies!show more

Buzz is Slack where the agents are teammates instead... of integrations. Here's what that unlocks: Agents: - The ones you start with are just defaults - Edit them, add your own, write their instructions like a system prompt Swappable Harness: - This is the killer feature - Swap Claude Code, Codex, Goose under the same agent Context: - Every chat you've had with that agent carries over to the new harness - Global skills and skill files come too Think: Slack = the shape Agents = the teammates Harness = the engine you can change without losing anything "I get model fatigue" is a real thing. Buzz sits one layer above it.show more

🚨NEW EPISODE: AI Is Changing How We Work with... Crypto Data🚨 In today's episode, we're joined by Kofi to discuss how AI agents are fundamentally changing crypto data analysis, research, and engineering workflows. We discuss: - The evolution from manual coding to AI-assisted development - How MCPs unlock onchain data exploration - Using Herd MCP to analyze smart contracts - AI agents for protocol research and data engineering - Context windows, planning, and agent workflows - Cursor vs OpenCode vs Claude Code (Anthropic) - Managing multiple agents simultaneously - Model selection, costs, and productivity tradeoffs - Building internal AI knowledge bases - Agent orchestration and autonomous workflows - Training agents to replicate human expertise - The future of AI-powered crypto analytics - And much more—enjoy! — Timestamps: (00:00) Introduction (01:24) AI replaces frontends (03:24) Herd MCP demo (05:40) Exploring contracts with AI (08:06) Exploration, planning, execution (11:38) Investigating onchain activity (17:10) Managing agent workflows (20:20) Multi-agent productivity (23:01) Model selection tradeoffs (29:11) Trusting AI outputs (31:28) Missing links today (34:37) Agent orchestration future (38:02) Training agents from work (38:20) Outroshow more

HERMES JUST FIXED THE BIGGEST PROBLEM WITH BROWSER AGENTS... Most AI agents still click around websites like confused interns. This new setup gives them the map. Hermes + → Hermes now connects to Browserbase’s new Browse hub → Browse launched with 100+ browser skills → Each skill is a plain-text playbook for a specific website or task → Your agent can search, preview, and install skills inside Hermes Why This Matters: ✓ Less random clicking ✓ Fewer timeouts ✓ Better form filling ✓ More reliable website navigation ✓ Skills can be edited, reused, bundled, and shared The Real Stack: → Hermes runs the agent → Browserbase handles cloud browser infrastructure → gives the agent site-specific skills → Vision helps when pages get weird → Bundles let you load repeat workflows faster The killer detail: If a website breaks, you don’t wait for the model to magically improve. You update the skill. Now your agent gets better forever. That’s the difference between an AI tool and an actual agent system.show more

Support for Agent Plugins is live in Cursor (and... I made a plugin for searching X!) Agent Plugins combine Skills and MCP servers into a single package Think about them like an Ikea furniture kit: your table comes with instructions, parts, and tools. You use the instructions to understand how things work (which tools on which parts), then use the tools to assemble your furniture Agent Plugins are bundles of tools & resources that allow agents to accomplish a goal: - MCP servers: fixed context and tools - Skills: dynamic instructions, loaded progressively Plugins bundle tools and instructions so your agent is able to use tokens efficiently Many of us already use MCP servers and skills together - Agent Plugins formalize this pattern for more powerful workflows I'm excited to see where the standard takes us!show more

How will companies build agents that work alongside their... employees? I discuss this with Guillermo Rauch, the CEO of Vercel, who’s building V... an internal agent used by almost 1,000 people across Vercel. We talk about: - The Openclaw Movement - How to get started building agents for your team? - Do we build a team of many agents or one GOD agent? - Hard parts of creating agents - Agents that delegate to other agents with different permissions and computer access - Eve The framework for building agents This is Agent Native #4 00:00 Intro 01:21 The State of Agent Adoption 07:26 "V" - Vercel's Internal Company Agent 14:32 Lessons from OpenClaw: Soul & Computer use 22:20 One God Agent OR a Team of Agents 27:11 Build Your Own Agent: Skills, Tools & Permissions 33:52 Proactive Agents, Triggers & Self-Improvement 40:30 Who Will Build Agents & Getting Started 46:22 Open Models - Cheaper, Faster, Better 53:05 Closing Thoughts: Owning Your Own Intelligenceshow more

New Course: Post-training of LLMs Learn to post-train and... customize an LLM in this short course, taught by Banghua Zhu, Assistant Professor at the University of Washington University of Washington, and co-founder of @NexusflowX. Training an LLM to follow instructions or answer questions has two key stages: pre-training and post-training. In pre-training, it learns to predict the next word or token from large amounts of unlabeled text. In post-training, it learns useful behaviors such as following instructions, tool use, and reasoning. Post-training transforms a general-purpose token predictor—trained on trillions of unlabeled text tokens—into an assistant that follows instructions and performs specific tasks. Because it is much cheaper than pre-training, it is practical for many more teams to incorporate post-training methods into their workflows than pre-training. In this course, you’ll learn three common post-training methods—Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Online Reinforcement Learning (RL)—and how to use each one effectively. With SFT, you train the model on pairs of input and ideal output responses. With DPO, you provide both a preferred (chosen) and a less preferred (rejected) response and train the model to favor the preferred output. With RL, the model generates an output, receives a reward score based on human or automated feedback, and updates the model to improve performance. You’ll learn the basic concepts, common use cases, and principles for curating high-quality data for effective training. Through hands-on labs, you’ll download a pre-trained model from Hugging Face and post-train it using SFT, DPO, and RL to see how each technique shapes model behavior. In detail, you’ll: - Understand what post-training is, when to use it, and how it differs from pre-training. - Build an SFT pipeline to turn a base model into an instruct model. - Explore how DPO reshapes behavior by minimizing contrastive loss—penalizing poor responses and reinforcing preferred ones. - Implement a DPO pipeline to change the identity of a chat assistant. - Learn online RL methods such as Proximal Policy Optimization (PPO) and Group Relative Policy Optimization (GRPO), and how to design reward functions. - Train a model with GRPO to improve its math capabilities using a verifiable reward. Post-training is one of the most rapidly developing areas of LLM training. Whether you’re building a high-accuracy context-specific assistant, fine-tuning a model's tone, or improving task-specific accuracy, this course will give you experience with the most important techniques shaping how LLMs are post-trained today. Please sign up here:show more

Google just dropped a free 1-hour course On building... complete agent systems from scratch: 07:02 - Build your first AI agent 13:34 - Give every agent a specialized skill 23:08 - Engineer the right context 32:56 - Connect agents into graphs 54:11 - Run 1,000 agents in autonomous loops One agent saves you an hour A thousand agents replace the team you cannot afford to hire Most people are still learning how to prompt one chatbot Google is already teaching how to build entire agent systems Agent → Skills → Context → Graphs → Loops Single agents are the old workflow Agent swarms are the new one This free 1-hour course is worth more than most paid agent engineering programs Bookmark it before everyone starts building the same thing Then build your first agent with the guide belowshow more

1/3 Money, money, money, moneyyyyy 💸💸💸💸 Today we’re making... it possible for you to earn actual money from your Pika AI Self agent. Because we think your agent should work FOR you in every sense of the phrase. Every time someone talks with them, or uses one of their skills, you earn tokens redeemable for cash. Say goodbye to those deadbeat agents.show more

OpenAI just launched "agent skills", their version of what... Claude calls skills. It finally clicked to me how these agents, skills and MCPs will work together. Basically, a skill is just the recipe for how something should be done, subagents are extra hands you spin up when the task is big, and MCPs are how the agent actually touches real systems like your repo or dashboards. Imagine a solo graphic designer setting up a “launch assets” skill: the skill defines brand rules, layouts, and export formats, sub-agents generate variations for social, web, and email in parallel, and MCPs pull from the actual Figma library and push finished assets into the client’s folder. Cool to see this kind of structure emerge in both OpenAI and Claude.... it feels like an early step toward a common way of building workflows that are reliable, composable, and actually useful in real work. I also gave away a trend, a business idea, an app ive been using for 14 years and more. (i answer every comment there) Happy building, my friends. Btw, what should I cover next? Or tag who should come on the show. Have a creative day.show more