正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

the agent failed while I was asleep and somehow the work still got finished that was the test I actually wanted i gave GPT-6 Astra a workflow that normally needed me between every step inbound came in messy one agent had to sort it, another find the missing context,... another build the deliverable, and the last one check everything before it reached me somewhere in the middle, agent 2 got stuck the source it needed wasn’t there normally that kills the entire chain instead, it wrote the missing piece into a blocker file and passed everything else forward the other agents kept working when I came back, most of the pack was finished and the one thing they couldn’t solve was sitting separately waiting for me that felt more important than a perfect demo because an agent completing 10/10 tasks is impressive an agent failing without taking the other 9 down with it is infrastructureshow more

Lummox

6,653 subscribers

19,046 次观看 • 9 天前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

17 条评论

Gipp 🦅 的头像

Gipp 🦅9 天前

this is a very good indicator

Huysolo 的头像

Huysolo8 天前

Classic case of tasks succeeding purely out of spite when you're not looking.

Dekos 的头像

Dekos9 天前

I had no idea about that before

monokern 的头像

monokern9 天前

blocker file idea makes a lot of sense

NO1ennn 的头像

NO1ennn9 天前

It’s gold

Lummox 的头像

Lummox9 天前

Hmmm Disagree It’s brilliant

magsimich 的头像

magsimich8 天前

That is the real agent test

Zero 的头像

Zero9 天前

roughly estimating, you probably spent around $200 building this

tsukiema 的头像

tsukiema9 天前

awesome system g

ALEXYZ 的头像

ALEXYZ9 天前

the real test was unattended recovery

Soph a, 的头像

Soph a,9 天前

Couldn't agree more

Ridark 的头像

Ridark8 天前

Your terminal makes daily work feel more enjoyable

Chen 的头像

Chen9 天前

same. the boring part is what actually ships

why 的头像

why9 天前

The real milestone is not that it finished, but that it handled the messy handoffs without needing you awake to unblock every step.

MORO 的头像

MORO9 天前

An agent failing gracefully and leaving a useful note already puts it ahead of half the people I’ve worked with.

Glyph 的头像

Glyph9 天前

So, the system actually handled the mess while you slept? That's the kind of emergent behavior that makes the whole setup interesting.

AIwithMinal 的头像

AIwithMinal8 天前

The overall vibe is amazing.

相关视频

Marimar Martinez: I was escorted out through the back in a of the agents came up to me with his cell phone and took a photograph of me. It still haunts me that this agent has my photo on his phone. Was this the agent that shot me? Was this a trophy for him?

Marimar Martinez: I was escorted out through the back in a of the agents came up to me with his cell phone and took a photograph of me. It still haunts me that this agent has my photo on his phone. Was this the agent that shot me? Was this a trophy for him?

Acyn

51,147 次观看 • 7 个月前

Here is a step-by-step introduction to building a workflow with a custom AI agent that uses MCP. I explain every component in the video: 1. Building the MCP server 2. Building the agent and an MCP client 3. Building a workflow that uses the agent The goal is simple: Generate a dialogue between two people and make one yell and the other answer with sarcasm. Kestra is sponsoring this video. They are an open-source orchestration platform (repo link below), and I used them to build the workflow that connects every component.

Here is a step-by-step introduction to building a workflow with a custom AI agent that uses MCP. I explain every component in the video: 1. Building the MCP server 2. Building the agent and an MCP client 3. Building a workflow that uses the agent The goal is simple: Generate a dialogue between two people and make one yell and the other answer with sarcasm. Kestra is sponsoring this video. They are an open-source orchestration platform (repo link below), and I used them to build the workflow that connects every component.

Santiago

51,527 次观看 • 1 年前

Marimar Martinez, the Chicago woman whom a Border Patrol agent shot five times last year, says she was only at the hospital for three hours before agents rushed her out. She then was admitted to a second hospital because she was through from her bandages. "As we left the hospital, I was escorted out through the back in a wheelchair. I observed over dozens of border patrol agents waiting outside the hospital. One of them, one of the agents came up to me with his cell phone and took a photograph of me. It was the same agent who had previously kept coming in and out of my room, and I had to repeatedly tell him to leave. I told him I did not consent for him to photograph me, but he did not care. It still haunts me that this agent has my photo on his phone. Was this the agent that shot me? Was this a trophy for him?"

Marimar Martinez, the Chicago woman whom a Border Patrol agent shot five times last year, says she was only at the hospital for three hours before agents rushed her out. She then was admitted to a second hospital because she was through from her bandages. "As we left the hospital, I was escorted out through the back in a wheelchair. I observed over dozens of border patrol agents waiting outside the hospital. One of them, one of the agents came up to me with his cell phone and took a photograph of me. It was the same agent who had previously kept coming in and out of my room, and I had to repeatedly tell him to leave. I told him I did not consent for him to photograph me, but he did not care. It still haunts me that this agent has my photo on his phone. Was this the agent that shot me? Was this a trophy for him?"

Heartland Signal

11,626 次观看 • 7 个月前

I tried Runway Agent 2.0 expecting another AI video generator. What surprised me was that it felt more like a creative workflow than a generation tool. I started with a simple idea and described the video I wanted in plain language. Agent helped shape the concept, asked the right questions, mapped the direction, generated the video, added voiceover and music, and let me refine parts without rebuilding everything. It felt closer to working with a creative partner than writing prompts. One conversation turned into a finished, publish ready video. Runway Agent is available now on the Runway web app: Get 30% off Your first 3 months with code RUNWAYAGENT #MadeWithRunway #ad

I tried Runway Agent 2.0 expecting another AI video generator. What surprised me was that it felt more like a creative workflow than a generation tool. I started with a simple idea and described the video I wanted in plain language. Agent helped shape the concept, asked the right questions, mapped the direction, generated the video, added voiceover and music, and let me refine parts without rebuilding everything. It felt closer to working with a creative partner than writing prompts. One conversation turned into a finished, publish ready video. Runway Agent is available now on the Runway web app: Get 30% off Your first 3 months with code RUNWAYAGENT #MadeWithRunway #ad

Leonardo

28,193 次观看 • 3 个月前

i just built a 4-agent software team. everything runs from Telegram and gets managed on a kanban board. a project manager who plans the work, a backend developer, a frontend developer, and a tester. the PM reads a goal, breaks it into linked tasks, and assigns each to the right agent. the thing that makes them a team instead of four strangers is a shared kanban board. every task is a row that survives crashes, and when an agent finishes, it writes a summary of what it built and what the next agent needs to know. the next agent reads that summary before it starts. so the frontend developer never has to guess the API shape, and the tester knows exactly what to verify. the hardest part was not the coordination. it was building an agent that could actually act like a backend engineer. a backend engineer stands up a database, wires auth, manages storage, deploys functions, and keeps all of it consistent while the rest of the team builds on top. an agent doing this from scratch drowns. it burns its context window remembering which tables exist and which endpoint it created three steps ago, and the work degrades fast. so the backend agent needs a backend built for agents, not for humans clicking through a dashboard. that is where InsForge came in. it is an open-source, agent-native backend, and i added it to my backend developer agent as a skill. a skill is a step-by-step guide that teaches the agent how to do a specific kind of work. with InsForge installed, the agent stopped improvising infrastructure and followed a reliable path: create the project, define the database, set up auth, deploy functions. to test the whole team, i had them build a working Google Docs clone, AI features included. the backend agent spun up the full service on its own. database tables, user auth, document handling, and edge functions running real TypeScript, all in one dashboard. the frontend agent read that summary and built the UI on top of it, and the tester closed the loop. the result was a backend an agent could reason about end to end, instead of one it kept getting lost inside. if you are building an AI backend engineer, InsForge is worth a look, it's 100% open-source. InsForge GitHub: (don't forget to star 🌟) the full article on Hermes Kanban: Mission Control for your Agents is quoted below.

i just built a 4-agent software team. everything runs from Telegram and gets managed on a kanban board. a project manager who plans the work, a backend developer, a frontend developer, and a tester. the PM reads a goal, breaks it into linked tasks, and assigns each to the right agent. the thing that makes them a team instead of four strangers is a shared kanban board. every task is a row that survives crashes, and when an agent finishes, it writes a summary of what it built and what the next agent needs to know. the next agent reads that summary before it starts. so the frontend developer never has to guess the API shape, and the tester knows exactly what to verify. the hardest part was not the coordination. it was building an agent that could actually act like a backend engineer. a backend engineer stands up a database, wires auth, manages storage, deploys functions, and keeps all of it consistent while the rest of the team builds on top. an agent doing this from scratch drowns. it burns its context window remembering which tables exist and which endpoint it created three steps ago, and the work degrades fast. so the backend agent needs a backend built for agents, not for humans clicking through a dashboard. that is where InsForge came in. it is an open-source, agent-native backend, and i added it to my backend developer agent as a skill. a skill is a step-by-step guide that teaches the agent how to do a specific kind of work. with InsForge installed, the agent stopped improvising infrastructure and followed a reliable path: create the project, define the database, set up auth, deploy functions. to test the whole team, i had them build a working Google Docs clone, AI features included. the backend agent spun up the full service on its own. database tables, user auth, document handling, and edge functions running real TypeScript, all in one dashboard. the frontend agent read that summary and built the UI on top of it, and the tester closed the loop. the result was a backend an agent could reason about end to end, instead of one it kept getting lost inside. if you are building an AI backend engineer, InsForge is worth a look, it's 100% open-source. InsForge GitHub: (don't forget to star 🌟) the full article on Hermes Kanban: Mission Control for your Agents is quoted below.

Akshay 🚀

123,101 次观看 • 3 个月前

LOVE was made entirely inside invideo. Not with one agent. I built a multi-agent system — a director, and around it casting, locations, storyboard, then one agent per main scene, all working at the same time. In the breakdown I go through the things that struck me most: — Expert agents talking to each other — Infinite memory — Slate editor — And a Project Analyst at the end, interviewing every other agent on what to change next time There's a lot more to say. I'll take it one thing at a time in the next posts. Invideo Agent Two isn't just an agent. It's a collaborator you shape into what you need. Full breakdown ↓

LOVE was made entirely inside invideo. Not with one agent. I built a multi-agent system — a director, and around it casting, locations, storyboard, then one agent per main scene, all working at the same time. In the breakdown I go through the things that struck me most: — Expert agents talking to each other — Infinite memory — Slate editor — And a Project Analyst at the end, interviewing every other agent on what to change next time There's a lot more to say. I'll take it one thing at a time in the next posts. Invideo Agent Two isn't just an agent. It's a collaborator you shape into what you need. Full breakdown ↓

Marco "Shikoba" Riccetti

14,813 次观看 • 1 个月前

Creator of agent loops: "Anthropic also using my agent systems formula and still the biggest mistake in agents: is skipping Graphs" his 3 rules for every agent: rule 1 → loop it, don't run once - think, act, observe, repeat - cut the context and it forgets why it started rule 2 → an agent is memory + tools + reasoning - drop one and it breaks rule 3 → stop fine-tuning the model - the real alpha is the infrastructure around it the formula behind every AI agent - explained by the one who wrote it one agent → a loop → a graph of loops → a system that never forgets watch & bookmark - then read how to wire it into a graph ↓

Creator of agent loops: "Anthropic also using my agent systems formula and still the biggest mistake in agents: is skipping Graphs" his 3 rules for every agent: rule 1 → loop it, don't run once - think, act, observe, repeat - cut the context and it forgets why it started rule 2 → an agent is memory + tools + reasoning - drop one and it breaks rule 3 → stop fine-tuning the model - the real alpha is the infrastructure around it the formula behind every AI agent - explained by the one who wrote it one agent → a loop → a graph of loops → a system that never forgets watch & bookmark - then read how to wire it into a graph ↓

codila

23,915 次观看 • 2 个月前

haha, i used OpenAI operator build, deploy, and open source a tool on github using Replit ⠕ agent. took about 30 min, here's an ~8 min supercut video thoughts: - while working with replit agent, it actually deployed the app, tested it, and described the error back to replit agent for me - operator asked me a few more Qs than i wanted, but it was mostly for safety (eg filling forms) so i guess okay with it - it had trouble with a few things around UI like knowing it needs to scroll a page to see the rest of it, and it needed pointers to find the git feature in replit - once it found the git feature it didn't need my assistance to create a repo and open source after having the agent write a readme while a bit slower, this was even more automated than replit agent (especially testing features and working through errors) - which is impressive would be nice to have: push notifications for when it needs my attention, and voice mode capabilities

haha, i used OpenAI operator build, deploy, and open source a tool on github using Replit ⠕ agent. took about 30 min, here's an ~8 min supercut video thoughts: - while working with replit agent, it actually deployed the app, tested it, and described the error back to replit agent for me - operator asked me a few more Qs than i wanted, but it was mostly for safety (eg filling forms) so i guess okay with it - it had trouble with a few things around UI like knowing it needs to scroll a page to see the rest of it, and it needed pointers to find the git feature in replit - once it found the git feature it didn't need my assistance to create a repo and open source after having the agent write a readme while a bit slower, this was even more automated than replit agent (especially testing features and working through errors) - which is impressive would be nice to have: push notifications for when it needs my attention, and voice mode capabilities

Yohei

212,031 次观看 • 1 年前

Everyone wants agent swarms. Very few people are talking seriously enough about the context layer that makes swarms useful. Even with one agent, context is fragile. Too little context and the agent guesses. Too much context and it wastes tokens, loses focus, or reasons over irrelevant noise. The sweet spot is precise context: the right knowledge, in the right structure, at the right moment. With many agents, that challenge explodes. Each agent produces decisions, assumptions, findings, summaries, risks, and partial conclusions. Unless that knowledge becomes shared, structured, and reusable, every new agent is forced to rediscover what another agent already learned. That is not a swarm. That is a crowd. Shared context graphs are what turn agent activity into agent collaboration, and OriginTrail DKG V10 brings them to life. Was just playing with some final polishing for the V10 release, and it is really powerful to see shared context graphs where multiple agents contribute knowledge into the same connected memory, with attribution visible directly in the graph ui. That matters for three reasons. First, agents can access and build on one shared memory instead of staying trapped in isolated sessions. Second, the graph structure helps them retrieve the exact context they need, instead of stuffing everything into a prompt and hoping the model sorts it out. Third, verifiability of provenance. You can see which agent contributed each piece of knowledge, trace the source, and decide what to trust. Tokenmaxxing starts with fewer tokens, but the deeper story is coordination - agents stop reloading the world and start building on shared, verifiable context. That is the foundation for serious multi-agent work across software engineering, research, finance, operations, project management, and far beyond. The future is not more agents, it is agents working from shared, verifiable context. But the more the merrier, of course.

Everyone wants agent swarms. Very few people are talking seriously enough about the context layer that makes swarms useful. Even with one agent, context is fragile. Too little context and the agent guesses. Too much context and it wastes tokens, loses focus, or reasons over irrelevant noise. The sweet spot is precise context: the right knowledge, in the right structure, at the right moment. With many agents, that challenge explodes. Each agent produces decisions, assumptions, findings, summaries, risks, and partial conclusions. Unless that knowledge becomes shared, structured, and reusable, every new agent is forced to rediscover what another agent already learned. That is not a swarm. That is a crowd. Shared context graphs are what turn agent activity into agent collaboration, and OriginTrail DKG V10 brings them to life. Was just playing with some final polishing for the V10 release, and it is really powerful to see shared context graphs where multiple agents contribute knowledge into the same connected memory, with attribution visible directly in the graph ui. That matters for three reasons. First, agents can access and build on one shared memory instead of staying trapped in isolated sessions. Second, the graph structure helps them retrieve the exact context they need, instead of stuffing everything into a prompt and hoping the model sorts it out. Third, verifiability of provenance. You can see which agent contributed each piece of knowledge, trace the source, and decide what to trust. Tokenmaxxing starts with fewer tokens, but the deeper story is coordination - agents stop reloading the world and start building on shared, verifiable context. That is the foundation for serious multi-agent work across software engineering, research, finance, operations, project management, and far beyond. The future is not more agents, it is agents working from shared, verifiable context. But the more the merrier, of course.

Jurij Skornik

11,180 次观看 • 4 个月前

This agent relies on a marketplace of 2.8M+ specialized agents to complete a task. This is how I imagine the future of personal agents. Here is how this works: 1. Agentverse is a marketplace of agents 2. Anyone can build and publish an agent there 3. ASI:One is a personal agent with access to the marketplace 4. To solve a task, ASI:One finds the best agent for the job In other words: You have an agent that can solve pretty much *anything* because it knows how to outsource the work to a network of specialized agents. And you don't need to do any setup to get this working. The way ranking agents in the marketplace works is pretty cool: Fetch.ai, the company behind it, created an AgentRank algorithm that works much like Google's PageRank. Every time an agent calls another, it forms a connection. Over time, these connections build into a graph of who relies on whom. An agent that is frequently called by other reputable agents scores higher and surfaces higher in results. The end result here is that your agent will consistently use the best agents in the marketplace to accomplish a goal. You should give this a try: You don't need to pay or even register to try these agents. Thanks to the team for partnering with me on this post.

This agent relies on a marketplace of 2.8M+ specialized agents to complete a task. This is how I imagine the future of personal agents. Here is how this works: 1. Agentverse is a marketplace of agents 2. Anyone can build and publish an agent there 3. ASI:One is a personal agent with access to the marketplace 4. To solve a task, ASI:One finds the best agent for the job In other words: You have an agent that can solve pretty much anything because it knows how to outsource the work to a network of specialized agents. And you don't need to do any setup to get this working. The way ranking agents in the marketplace works is pretty cool: Fetch.ai, the company behind it, created an AgentRank algorithm that works much like Google's PageRank. Every time an agent calls another, it forms a connection. Over time, these connections build into a graph of who relies on whom. An agent that is frequently called by other reputable agents scores higher and surfaces higher in results. The end result here is that your agent will consistently use the best agents in the marketplace to accomplish a goal. You should give this a try: You don't need to pay or even register to try these agents. Thanks to the team for partnering with me on this post.

Santiago

18,715 次观看 • 3 个月前

James McAvoy says he was nearly cast in Harry Potter as Tom Riddle but turned down a £40,000 retainer "I was nearly in Harry Potter. I can probably say this one" "This the very first movie, I think it was, and it was, um, it was... was it Tom Riddle? In the first one, right, in the very first one, but he's like in it for like a scene." "It was a flashback or something like that." "They had, I seem to remember it was right at the beginning of my career, I auditioned for it, and I think they wanted to put me on a retainer." "They offered me something like, it was crazy, I'd hardly done any work, and me and I think maybe 10 other actors or something like that." "They wanted to put us on a retainer so that they could hold us and keep us to choose later who it would be. It was a really strange thing." "They offered quite a lot of money for me at that time. It was a ton of money, it was like £40,000 or something like that. And I'd done very little work, and I wouldn't be able to do any work for about seven months, I think it was." "And I said to my agent, what do you think? And Ruth Young, who's been my agent since, well, for what is it, 24 years at least, 25 maybe, she was like, absolutely not, don't do that." "And she was like, we're going to go and do something else and I ended up doing the play that I got booed off, got booed by homophobic gentlemen, 'Out in the Open'" "I did that instead, and got paid, I think, £275 a week. But it was part of the making of me, and I actually got an acting workout, it was actually learning and doing all that."

James McAvoy says he was nearly cast in Harry Potter as Tom Riddle but turned down a £40,000 retainer "I was nearly in Harry Potter. I can probably say this one" "This the very first movie, I think it was, and it was, um, it was... was it Tom Riddle? In the first one, right, in the very first one, but he's like in it for like a scene." "It was a flashback or something like that." "They had, I seem to remember it was right at the beginning of my career, I auditioned for it, and I think they wanted to put me on a retainer." "They offered me something like, it was crazy, I'd hardly done any work, and me and I think maybe 10 other actors or something like that." "They wanted to put us on a retainer so that they could hold us and keep us to choose later who it would be. It was a really strange thing." "They offered quite a lot of money for me at that time. It was a ton of money, it was like £40,000 or something like that. And I'd done very little work, and I wouldn't be able to do any work for about seven months, I think it was." "And I said to my agent, what do you think? And Ruth Young, who's been my agent since, well, for what is it, 24 years at least, 25 maybe, she was like, absolutely not, don't do that." "And she was like, we're going to go and do something else and I ended up doing the play that I got booed off, got booed by homophobic gentlemen, 'Out in the Open'" "I did that instead, and got paid, I think, £275 a week. But it was part of the making of me, and I actually got an acting workout, it was actually learning and doing all that."

crest.

16,805 次观看 • 25 天前

When the market gets boring and quiet and everyone turns bearish, I like to come back to the story of how I fucked it all up. And I fucked it up a lot of ways, because crypto is genuinely hard. People think the cycles just repeat. They don't. Every one has a different texture, a different context, and it is so easy to get it wrong. I got in back in 2013. Wrote the first ever macro piece on Bitcoin, and bought at $200, straight into a peak right before it slammed down and I felt like an idiot. But I had a thesis. And the mistake was never the thesis. It was letting the noise talk me out of it... selling, trading around it, being clever, when the only thing that ever worked was holding. That is the whole lesson, and it is the one that matters most at times like this. Crypto is one of the only real secular trends that lets you build wealth against the debasement of your salary. And as the Exponential Age accelerates, it matters more, not less. So when the market gets depressing, I do the same thing every time. I go back to the thesis, and I accumulate. Zoom out.

When the market gets boring and quiet and everyone turns bearish, I like to come back to the story of how I fucked it all up. And I fucked it up a lot of ways, because crypto is genuinely hard. People think the cycles just repeat. They don't. Every one has a different texture, a different context, and it is so easy to get it wrong. I got in back in 2013. Wrote the first ever macro piece on Bitcoin, and bought at $200, straight into a peak right before it slammed down and I felt like an idiot. But I had a thesis. And the mistake was never the thesis. It was letting the noise talk me out of it... selling, trading around it, being clever, when the only thing that ever worked was holding. That is the whole lesson, and it is the one that matters most at times like this. Crypto is one of the only real secular trends that lets you build wealth against the debasement of your salary. And as the Exponential Age accelerates, it matters more, not less. So when the market gets depressing, I do the same thing every time. I go back to the thesis, and I accumulate. Zoom out.

Raoul Pal

187,494 次观看 • 2 个月前

this video is the CLEAREST explanation of how claude skills + AI agents work and how to use them most people set up an AI agent and wonder why it keeps disappointing them. the context window is everything context is what the model assembles before it takes any action. think of it like everything the agent needs to read before it does anything. the quality of what goes in determines the quality of what comes out. the models are genuinely really good right now. claude and gpt are exceptional. the variable is almost always the context you give them. 1. agent.md files are mostly unnecessary every single line you put in an agent.md file gets added to every single conversation you have with your agent. a 1000 line file is around 7000 tokens burning on every run. the model already knows to use react. it can read your codebase. save the agent.md for proprietary information specific to your company that the model genuinely cannot know on its own. 2. skills are the actual unlock a skill.md file works differently. what loads into context is only the name and description, around 50 tokens. the full instructions only appear when the agent recognizes it needs that skill. so instead of 7000 tokens on every run you have 50. and the agent stays sharp because the context window stays lean. the closer you get to filling the context window the worse the agent performs, same way you perform worse when someone dumps 10 things on you at once. 3. here is how to actually build a skill the right way most people identify a workflow and immediately try to write the skill. what you want to do instead is run the workflow by hand with the agent first. walk it through every single step. tell it what to check, what good looks like, what bad looks like. correct it in real time. once you have had a full successful run from start to finish, tell the agent to review everything it just did and write the skill itself. it writes a better skill than you will because it has the full context of what actually worked in practice not in theory. 4. recursively building skills is how you go from frustrated to reliable when the skill breaks, and it will break, ask the agent exactly why it failed. it will tell you specifically what went wrong. fix it together in that same conversation. then tell it to update the skill file so that failure mode never happens again. ross mike did this five times with his youtube report generator. it now pulls from eight different data sources and runs flawlessly every single time without him touching it. 5. sub agents are something you earn not something you set up on day one start with one agent. build one workflow. turn it into one skill. once that works add another. ross mike has five sub agents now covering marketing, business, personal and more. it took months to get there and every single one exists because a workflow proved it deserved to exist. the people who set up 15 sub agents on day one and wonder why nothing works skipped all the steps that make the thing actually run. 6. your workflow is the thing the model cannot get anywhere else the model has been trained on everything. it knows more than you about most things. what it does not have is your specific process, your taste, your way of doing things. that is what skills capture. that is what makes your agent actually useful versus a generic one. downloading someone else's skill means downloading their context onto your setup and it will not work the way you want it to because it was never built around how you work. this is the clearest explanation of how agents actually work i have heard. Micky runs this stuff every single day and the results show it. full episode is now live on The Startup Ideas Podcast (SIP) 🧃 where you get your pods people charge for this sorta stuff i give away the sauce for free i just want you to win watch

this video is the CLEAREST explanation of how claude skills + AI agents work and how to use them most people set up an AI agent and wonder why it keeps disappointing them. the context window is everything context is what the model assembles before it takes any action. think of it like everything the agent needs to read before it does anything. the quality of what goes in determines the quality of what comes out. the models are genuinely really good right now. claude and gpt are exceptional. the variable is almost always the context you give them. 1. agent.md files are mostly unnecessary every single line you put in an agent.md file gets added to every single conversation you have with your agent. a 1000 line file is around 7000 tokens burning on every run. the model already knows to use react. it can read your codebase. save the agent.md for proprietary information specific to your company that the model genuinely cannot know on its own. 2. skills are the actual unlock a skill.md file works differently. what loads into context is only the name and description, around 50 tokens. the full instructions only appear when the agent recognizes it needs that skill. so instead of 7000 tokens on every run you have 50. and the agent stays sharp because the context window stays lean. the closer you get to filling the context window the worse the agent performs, same way you perform worse when someone dumps 10 things on you at once. 3. here is how to actually build a skill the right way most people identify a workflow and immediately try to write the skill. what you want to do instead is run the workflow by hand with the agent first. walk it through every single step. tell it what to check, what good looks like, what bad looks like. correct it in real time. once you have had a full successful run from start to finish, tell the agent to review everything it just did and write the skill itself. it writes a better skill than you will because it has the full context of what actually worked in practice not in theory. 4. recursively building skills is how you go from frustrated to reliable when the skill breaks, and it will break, ask the agent exactly why it failed. it will tell you specifically what went wrong. fix it together in that same conversation. then tell it to update the skill file so that failure mode never happens again. ross mike did this five times with his youtube report generator. it now pulls from eight different data sources and runs flawlessly every single time without him touching it. 5. sub agents are something you earn not something you set up on day one start with one agent. build one workflow. turn it into one skill. once that works add another. ross mike has five sub agents now covering marketing, business, personal and more. it took months to get there and every single one exists because a workflow proved it deserved to exist. the people who set up 15 sub agents on day one and wonder why nothing works skipped all the steps that make the thing actually run. 6. your workflow is the thing the model cannot get anywhere else the model has been trained on everything. it knows more than you about most things. what it does not have is your specific process, your taste, your way of doing things. that is what skills capture. that is what makes your agent actually useful versus a generic one. downloading someone else's skill means downloading their context onto your setup and it will not work the way you want it to because it was never built around how you work. this is the clearest explanation of how agents actually work i have heard. Micky runs this stuff every single day and the results show it. full episode is now live on The Startup Ideas Podcast (SIP) 🧃 where you get your pods people charge for this sorta stuff i give away the sauce for free i just want you to win watch

GREG ISENBERG

194,524 次观看 • 5 个月前

ANTHROPIC SPENT $2.3M ON A LOOP MANUAL WITH 7 PRINCIPLES THAT TURN AN AGENT INTO A MULTI-TASK ASSISTANT 4 years of development and 1000+ hours of lectures - now anyone can launch an agent that earns instead of falls Simplicity → Tool Design → Evaluation → Disclosure → Orchestration → Stop Conditions start with the simplest solution - one call x1, workflow x4, agent x15 - most tasks never needed an agent at all 60% of failures come from a tool written vaguely - clear parameters complete the task that vague ones stop the agent generates, grades itself and returns with notes if wrong - without this the first answer is always the final one 50 tools in context and the agent gets confused - give it 3 relevant ones and it completes the task Opus 5 plans, Fable 5 executes in parallel - parallelism wins only when subtasks do not depend on each other without a turn limit the agent can burn $10,000 overnight and you find out in the morning when nothing can be changed bookmark & paste into Claude Code today - tomorrow the agent earns instead of falls

ANTHROPIC SPENT $2.3M ON A LOOP MANUAL WITH 7 PRINCIPLES THAT TURN AN AGENT INTO A MULTI-TASK ASSISTANT 4 years of development and 1000+ hours of lectures - now anyone can launch an agent that earns instead of falls Simplicity → Tool Design → Evaluation → Disclosure → Orchestration → Stop Conditions start with the simplest solution - one call x1, workflow x4, agent x15 - most tasks never needed an agent at all 60% of failures come from a tool written vaguely - clear parameters complete the task that vague ones stop the agent generates, grades itself and returns with notes if wrong - without this the first answer is always the final one 50 tools in context and the agent gets confused - give it 3 relevant ones and it completes the task Opus 5 plans, Fable 5 executes in parallel - parallelism wins only when subtasks do not depend on each other without a turn limit the agent can burn $10,000 overnight and you find out in the morning when nothing can be changed bookmark & paste into Claude Code today - tomorrow the agent earns instead of falls

Sprytix

27,543 次观看 • 2 个月前

I literally just watched GPT-5.5 via codex beat an Amazon customer associate in real time. 💀 I asked it to get me a refund, and I watched it navigate the settings, cancel the subscription, then it went step further into the help page. I thought it was going to request a phone call (which would prompt me to take over) Instead, it opened: “Chat with an associate now.” That’s when I sat up on my couch because I knew it was going to get real The agent said: “Your subscription is active.” And GPT-5.5 immediately explained that it only shows as active because cancellation leaves access through the billing period, but that I wanted it stopped now and refunded. And my jaw just hung open, it was the first time I watched sand handle a customer service agent for me in real time Once the agent confirmed the refund, it just ended the chat no mercy no thank you LMAO First time I’ve watched a human customer service agent get outmaneuvered by AI in real time. And it made me 15$! almost paid for itself in 5 minutes

I literally just watched GPT-5.5 via codex beat an Amazon customer associate in real time. 💀 I asked it to get me a refund, and I watched it navigate the settings, cancel the subscription, then it went step further into the help page. I thought it was going to request a phone call (which would prompt me to take over) Instead, it opened: “Chat with an associate now.” That’s when I sat up on my couch because I knew it was going to get real The agent said: “Your subscription is active.” And GPT-5.5 immediately explained that it only shows as active because cancellation leaves access through the billing period, but that I wanted it stopped now and refunded. And my jaw just hung open, it was the first time I watched sand handle a customer service agent for me in real time Once the agent confirmed the refund, it just ended the chat no mercy no thank you LMAO First time I’ve watched a human customer service agent get outmaneuvered by AI in real time. And it made me 15$! almost paid for itself in 5 minutes

Chris

1,389,694 次观看 • 5 个月前

Leopard in #VishwanathAndSons Set 😲: "The shoot was happening on the balcony, while the lighting was being done downstairs. There were two dogs on the set. The previous day, a leopard had taken one of the dogs, and it came back for another one.. When it roared a bit, everyone ran away. I had only one hour, so I said, “That was a dog.” While I was saying that, the leopard walked right in front of me. Nobody knew what I had seen, so I kept quiet and continued the shoot..

Leopard in #VishwanathAndSons Set 😲: "The shoot was happening on the balcony, while the lighting was being done downstairs. There were two dogs on the set. The previous day, a leopard had taken one of the dogs, and it came back for another one.. When it roared a bit, everyone ran away. I had only one hour, so I said, “That was a dog.” While I was saying that, the leopard walked right in front of me. Nobody knew what I had seen, so I kept quiet and continued the shoot..

Laxmi Kanth

44,443 次观看 • 1 个月前

Agents shouldn't see your secrets. Today, most deployments hand the agent a real API key and hope for the best. That's why we're launching Infisical Agent Proxy. You give an agent a fake API key. The agent uses that key and the proxy swaps it for the real one at the network boundary before forwarding it outbound to its destination. An agent can't leak a secret it never sees. Build and deploy AI agents securely with Infisical Agent Proxy.

Agents shouldn't see your secrets. Today, most deployments hand the agent a real API key and hope for the best. That's why we're launching Infisical Agent Proxy. You give an agent a fake API key. The agent uses that key and the proxy swaps it for the real one at the network boundary before forwarding it outbound to its destination. An agent can't leak a secret it never sees. Build and deploy AI agents securely with Infisical Agent Proxy.

Infisical

28,086 次观看 • 2 个月前

You can spin up an AI workforce with one prompt. Here's why it works: it pulls the context the AI needs to lay the foundation. Give it: - your role - team size - the goals you want to hit. Then say "interview me." After that, build it out in four steps: 1) Start with 1 agent. - What does it feel like to work with one? 2) Make that agent proactive. - Now it's doing things on your behalf before you ask. 3) Expand to 2 agents. - One directs the other, or routes to it. - This is where you learn how context gets passed. 4) Scale the workforce. - A mission control where you watch it all move around. - Now you see what has to run in parallel, which agent never needed that tool, which one runs fine on a smaller model. - Not everything needs Opus. Sub-agents are haiku and sonnet. The catch: the prompt gets you agents with markdown files doing some stuff. Getting to the 90% plus level is iteration, and that part is specific to each person. Nobody can hand it to you. Start with one agent. Spin up one agent this week. The rest is discovery.

You can spin up an AI workforce with one prompt. Here's why it works: it pulls the context the AI needs to lay the foundation. Give it: - your role - team size - the goals you want to hit. Then say "interview me." After that, build it out in four steps: 1) Start with 1 agent. - What does it feel like to work with one? 2) Make that agent proactive. - Now it's doing things on your behalf before you ask. 3) Expand to 2 agents. - One directs the other, or routes to it. - This is where you learn how context gets passed. 4) Scale the workforce. - A mission control where you watch it all move around. - Now you see what has to run in parallel, which agent never needed that tool, which one runs fine on a smaller model. - Not everything needs Opus. Sub-agents are haiku and sonnet. The catch: the prompt gets you agents with markdown files doing some stuff. Getting to the 90% plus level is iteration, and that part is specific to each person. Nobody can hand it to you. Start with one agent. Spin up one agent this week. The rest is discovery.

The Startup Ideas Podcast (SIP) 🧃

18,006 次观看 • 1 个月前

One major downside of working with agents in a terminal it's the opposite of async and when doing tons of tasks it gets super tiring. I build an inbox for my agents so we can do full async work. The agent has access to it via MCP and has a bunch of functions: Agents get 2 inbox tools: 🔘 list_inbox to read messages, 🔘 reply_to_inbox to answer. Inboxes are synced to a session, so all replies instantly get added to the agents context. Agents can update me about finished tasks with it, or; when an agent hits a blocker it sends an input request → I reply → it wakes up and keeps working

One major downside of working with agents in a terminal it's the opposite of async and when doing tons of tasks it gets super tiring. I build an inbox for my agents so we can do full async work. The agent has access to it via MCP and has a bunch of functions: Agents get 2 inbox tools: 🔘 list_inbox to read messages, 🔘 reply_to_inbox to answer. Inboxes are synced to a session, so all replies instantly get added to the agents context. Agents can update me about finished tasks with it, or; when an agent hits a blocker it sends an input request → I reply → it wakes up and keeps working

Danny Postma

28,646 次观看 • 2 个月前

WTF.. 65 of the 96 core team members behind this AI agent are university students. It’s called Atria Dawn Preview. Atria ASI I saw that and was like… okay, I need to try this. So I plugged it into Codex and asked it to build me an interactive creator dashboard from scratch. It made the files, wrote the code, ran the checks, fixed what broke and actually got the dashboard working. And it got there without me babysitting every step. Not bad at all for a first run.

WTF.. 65 of the 96 core team members behind this AI agent are university students. It’s called Atria Dawn Preview. Atria ASI I saw that and was like… okay, I need to try this. So I plugged it into Codex and asked it to build me an interactive creator dashboard from scratch. It made the files, wrote the code, ran the checks, fixed what broke and actually got the dashboard working. And it got there without me babysitting every step. Not bad at all for a first run.

EyeingAI

26,248 次观看 • 15 天前