Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

“I’ll send notes after the call.” (They didn’t.) So we built an agent that does it automatically. Before a call, it pulls together the docs, decision log, and last week’s notes. After the meeting, it writes the recap and drops it in Slack. Pull up a chair. Sam H... show more

Notion

517,696 subscribers

16,474 Aufrufe • vor 4 Monaten •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

you should spend an afternoon building a second brain with jev... here's how i'd set it up: > create one inbox for saved posts, articles, meeting notes and ideas... give your agent access to it > give jev your active projects and interests so it can sort incoming material into categories that mean something to you > have your agent find similar notes, then use jev to flag duplicates and spot which ones add something new > save everything with its original link and date so you can always check where it came from > whenever you start a task, have your agent search the archive and let jev score which notes are relevant > feed those into your writing agent before it starts working... your customer conversations, examples and previous decisions are already there writing a proposal? bring back what the customer said on the call and the case study that answers their objection. working on a post? pull up the idea you saved three weeks ago alongside your own notes on the topic. start with one project and the material you've already collected for it. ask your coding agent to build that workflow, then check what it brings back on your next real task.

you should spend an afternoon building a second brain with jev... here's how i'd set it up: > create one inbox for saved posts, articles, meeting notes and ideas... give your agent access to it > give jev your active projects and interests so it can sort incoming material into categories that mean something to you > have your agent find similar notes, then use jev to flag duplicates and spot which ones add something new > save everything with its original link and date so you can always check where it came from > whenever you start a task, have your agent search the archive and let jev score which notes are relevant > feed those into your writing agent before it starts working... your customer conversations, examples and previous decisions are already there writing a proposal? bring back what the customer said on the call and the case study that answers their objection. working on a post? pull up the idea you saved three weeks ago alongside your own notes on the topic. start with one project and the material you've already collected for it. ask your coding agent to build that workflow, then check what it brings back on your next real task.

J.B.

13,511 Aufrufe • vor 12 Tagen

This is insane. An AI agent can run every boring job in outbound. We spent the last 6 months building ours. Here are my 8 favorite agents to build: Replies get sorted before we open the inbox. Campaigns go live from one command. Weak inboxes pull themselves out before they hurt a domain. Here are the agents behind it: 1. Reply Agent Reads every reply and drafts the response. A human reviews, edits, and sends. 2. Mailbox Health Agent Watches inbox and domain health. It predicts when you need new mailboxes, then buys and warms them. 3. Campaign Optimizer Agent Checks every live campaign every 6 hours. If 500+ leads were emailed and replies are under 4%, it tests new copy, replaces inboxes replying under 1%, and shifts sending to better hours. 4. Lead Qualification Agent Scores each lead by company size, industry, and tech stack. It enriches the record, updates your CRM, and only loads qualified leads into campaigns. High-priority prospect? You get a Slack ping. 5. Meeting Booking Agent Finds meeting requests inside replies. It books the slot, writes prep notes using the lead's background, sends reminders, and logs the outcome. 6. Pipeline Progression Agent Tracks opens, clicks, and website visits. It moves the CRM stage, triggers the next sequence, and creates a task when a lead shows real intent. 7. Copywriting Agent Writes cold emails and follow-ups in the campaign's voice. 8. Analytics Agent Watches campaign metrics in real time and explains what to fix in plain language. We built ours with custom code. Smartlead's SmartAgents let you build agents like these from a plain-English prompt, inside the platform where your campaigns already run. If you repeat an outbound task more than twice a week, that is an agent you have not built yet. Which one would you build first?

This is insane. An AI agent can run every boring job in outbound. We spent the last 6 months building ours. Here are my 8 favorite agents to build: Replies get sorted before we open the inbox. Campaigns go live from one command. Weak inboxes pull themselves out before they hurt a domain. Here are the agents behind it: 1. Reply Agent Reads every reply and drafts the response. A human reviews, edits, and sends. 2. Mailbox Health Agent Watches inbox and domain health. It predicts when you need new mailboxes, then buys and warms them. 3. Campaign Optimizer Agent Checks every live campaign every 6 hours. If 500+ leads were emailed and replies are under 4%, it tests new copy, replaces inboxes replying under 1%, and shifts sending to better hours. 4. Lead Qualification Agent Scores each lead by company size, industry, and tech stack. It enriches the record, updates your CRM, and only loads qualified leads into campaigns. High-priority prospect? You get a Slack ping. 5. Meeting Booking Agent Finds meeting requests inside replies. It books the slot, writes prep notes using the lead's background, sends reminders, and logs the outcome. 6. Pipeline Progression Agent Tracks opens, clicks, and website visits. It moves the CRM stage, triggers the next sequence, and creates a task when a lead shows real intent. 7. Copywriting Agent Writes cold emails and follow-ups in the campaign's voice. 8. Analytics Agent Watches campaign metrics in real time and explains what to fix in plain language. We built ours with custom code. Smartlead's SmartAgents let you build agents like these from a plain-English prompt, inside the platform where your campaigns already run. If you repeat an outbound task more than twice a week, that is an agent you have not built yet. Which one would you build first?

Hosun Chung

156,625 Aufrufe • vor 2 Monaten

Harness vs. Graphs, clearly explained! a harness is great, and most people think it is the whole thing: retries, timeouts, a sandbox, a log, the context it assembles before every call. all of that is real work, and all of it wraps exactly one call. run it a hundred times and you have one call, made very safely, a hundred times. Graph engineering fixes this by moving the decision up a layer: not how safely one call is made, but which calls exist to be made at all. you need both, and here is the sentence that resolves the whole confusion: the harness is everything around one call. the graph is everything between them. ↳ around one call: retry, timeout, sandbox, log, assemble the context, hand back a result ↳ between calls: split, fan out, merge, gate, send back Prompts → Context → Harness → Loops → Graphs the harness does not go away when you build a graph. it moves under each node, and now there are five of them, each wrapping a call you would never have made by hand. the trick is knowing which layer a failure belongs to. turn a piece off and run it again. if the call still works, it was the harness. if the wrong step runs at all, it was the graph. people spend weeks hardening a harness around a node that should not have existed. one thing to know before you scale it. most of what people call their agent is a harness with a chat box on it. ↳ it retries, it times out, it logs, it assembles context, it holds one call up beautifully ↳ it has never once decided that a second call should exist, and that is the entire difference that last one catches careful people. a harness that never fails is not evidence the system is right. it is evidence one call went well, which is the smallest possible claim. and the one that eats whole nights: a harness cannot save you from the wrong step running. you can retry a bad decision three times with a clean log and perfect isolation, and all you bought was three copies of it. below i have quoted my full guide on graph engineering. it covers the three topologies, the verifier patterns, and where the gate should actually open. save this and read it below ↓

Harness vs. Graphs, clearly explained! a harness is great, and most people think it is the whole thing: retries, timeouts, a sandbox, a log, the context it assembles before every call. all of that is real work, and all of it wraps exactly one call. run it a hundred times and you have one call, made very safely, a hundred times. Graph engineering fixes this by moving the decision up a layer: not how safely one call is made, but which calls exist to be made at all. you need both, and here is the sentence that resolves the whole confusion: the harness is everything around one call. the graph is everything between them. ↳ around one call: retry, timeout, sandbox, log, assemble the context, hand back a result ↳ between calls: split, fan out, merge, gate, send back Prompts → Context → Harness → Loops → Graphs the harness does not go away when you build a graph. it moves under each node, and now there are five of them, each wrapping a call you would never have made by hand. the trick is knowing which layer a failure belongs to. turn a piece off and run it again. if the call still works, it was the harness. if the wrong step runs at all, it was the graph. people spend weeks hardening a harness around a node that should not have existed. one thing to know before you scale it. most of what people call their agent is a harness with a chat box on it. ↳ it retries, it times out, it logs, it assembles context, it holds one call up beautifully ↳ it has never once decided that a second call should exist, and that is the entire difference that last one catches careful people. a harness that never fails is not evidence the system is right. it is evidence one call went well, which is the smallest possible claim. and the one that eats whole nights: a harness cannot save you from the wrong step running. you can retry a bad decision three times with a clean log and perfect isolation, and all you bought was three copies of it. below i have quoted my full guide on graph engineering. it covers the three topologies, the verifier patterns, and where the gate should actually open. save this and read it below ↓

Hanako

51,536 Aufrufe • vor 20 Tagen

For anyone constantly stuck in back-to-back meetings, Notion can automatically do all the follow ups. Just use AI meeting notes. Format a skill, hook it up to Notion mail, and every meeting after that handles itself.

For anyone constantly stuck in back-to-back meetings, Notion can automatically do all the follow ups. Just use AI meeting notes. Format a skill, hook it up to Notion mail, and every meeting after that handles itself.

Notion

46,431 Aufrufe • vor 5 Monaten

This choclift fan built an actual chocolate themed iPhone dock with a built-in NFC tag that automatically opens choclift the moment you place your iPhone on it. We call it docklift.

This choclift fan built an actual chocolate themed iPhone dock with a built-in NFC tag that automatically opens choclift the moment you place your iPhone on it. We call it docklift.

Phil Traut ᯅ

47,933 Aufrufe • vor 3 Monaten

THIS GUY CONNECTED HIS AI AGENTS TO HIS OBSIDIAN AND BUILT A BRAIN THAT LEARNS ON ITS OWN. HERE'S HOW TO BUILD IT Obsidian is just markdown files sitting in a folder. That turns out to be the perfect memory for an AI agent, because an agent can read and write those files directly. He wired his agents into the vault so they pull context from it, do the work, and write what they learned back. The notes aren't the point. The loop is, and it gets sharper every cycle How to build it: 1. Point an agent at your vault. The fastest way, no plugins, no API keys: open a terminal and run npx obsidian-mcp /path/to/your/vault. That exposes your Obsidian folder to Claude as a tool it can read, search, and write to. Add it to your Claude Code or Cowork config and restart 2. Confirm it can see the brain. Ask it: "list the notes in my vault and summarize what's in them." If it reads them back, the connection is live. Now it starts every task with everything the vault already holds instead of from zero 3. Give each agent one job and a write-back rule. Tell it: "research this, then save what you found as a new note in /brain with links to related notes." One agent researches, one summarizes, one plans. Each writes its output back into the vault 4. Close the loop. Add one line to every agent's instructions: "read /brain before starting, write your result back when done." Now each task leaves the vault richer, and the next run reads that before it works. It compounds instead of resetting 5. You only steer. Review what the brain produces, point it at the next thing. The agents handle the reading, writing, and connecting The edge isn't better notes. It's a brain that feeds itself, so the work gets sharper every cycle instead of starting over Bookmark this

THIS GUY CONNECTED HIS AI AGENTS TO HIS OBSIDIAN AND BUILT A BRAIN THAT LEARNS ON ITS OWN. HERE'S HOW TO BUILD IT Obsidian is just markdown files sitting in a folder. That turns out to be the perfect memory for an AI agent, because an agent can read and write those files directly. He wired his agents into the vault so they pull context from it, do the work, and write what they learned back. The notes aren't the point. The loop is, and it gets sharper every cycle How to build it: 1. Point an agent at your vault. The fastest way, no plugins, no API keys: open a terminal and run npx obsidian-mcp /path/to/your/vault. That exposes your Obsidian folder to Claude as a tool it can read, search, and write to. Add it to your Claude Code or Cowork config and restart 2. Confirm it can see the brain. Ask it: "list the notes in my vault and summarize what's in them." If it reads them back, the connection is live. Now it starts every task with everything the vault already holds instead of from zero 3. Give each agent one job and a write-back rule. Tell it: "research this, then save what you found as a new note in /brain with links to related notes." One agent researches, one summarizes, one plans. Each writes its output back into the vault 4. Close the loop. Add one line to every agent's instructions: "read /brain before starting, write your result back when done." Now each task leaves the vault richer, and the next run reads that before it works. It compounds instead of resetting 5. You only steer. Review what the brain produces, point it at the next thing. The agents handle the reading, writing, and connecting The edge isn't better notes. It's a brain that feeds itself, so the work gets sharper every cycle instead of starting over Bookmark this

Yarchi

58,591 Aufrufe • vor 3 Monaten

The reason your AI gives generic answers isn't the model. It's that your company's knowledge lives in 12 different tools and your agent can't see any of it. So here's how to build a context database your agents can actually use: 1) Pull it in: - Slack, email, meetings, docs, tickets - Auto-collected on a schedule into one place. 2) Clean it up. - An agent files what matters and drops the noise, so your brain doesn't bloat. 3) Store it as memory. - Folders of files your agents read from and write back to as they learn. 4) Point your agents at it. - Now they draft, build, and ship with full context instead of guessing. Same model. Same agent. The difference is whether it can understand your company or not. Build the database once. Everything downstream gets smarter.

The reason your AI gives generic answers isn't the model. It's that your company's knowledge lives in 12 different tools and your agent can't see any of it. So here's how to build a context database your agents can actually use: 1) Pull it in: - Slack, email, meetings, docs, tickets - Auto-collected on a schedule into one place. 2) Clean it up. - An agent files what matters and drops the noise, so your brain doesn't bloat. 3) Store it as memory. - Folders of files your agents read from and write back to as they learn. 4) Point your agents at it. - Now they draft, build, and ship with full context instead of guessing. Same model. Same agent. The difference is whether it can understand your company or not. Build the database once. Everything downstream gets smarter.

The Startup Ideas Podcast (SIP) 🧃

13,666 Aufrufe • vor 3 Monaten

A 19-year-old student built a little robot that goes to lectures and takes notes so they can sleep in class It is a tiny screen with an animated smiley face, sitting on a 3D-printed stickman body You just tap its face at the start of class It listens through a built-in mic, records the professor, and as soon as class ends, turns the entire lecture into a structured summary right on the display Under the hood, it runs on an ESP32 microcontroller, an OLED screen, a small battery, and an SD card The best part - the student did not even code it line by line. The entire codebase was written using Cursor AI Audio goes in, gets transcribed, AI structures the notes, and it is ready before you even leave the room Most students are still struggling to take notes while half asleep. This little guy just sits on the desk and handles it all

A 19-year-old student built a little robot that goes to lectures and takes notes so they can sleep in class It is a tiny screen with an animated smiley face, sitting on a 3D-printed stickman body You just tap its face at the start of class It listens through a built-in mic, records the professor, and as soon as class ends, turns the entire lecture into a structured summary right on the display Under the hood, it runs on an ESP32 microcontroller, an OLED screen, a small battery, and an SD card The best part - the student did not even code it line by line. The entire codebase was written using Cursor AI Audio goes in, gets transcribed, AI structures the notes, and it is ready before you even leave the room Most students are still struggling to take notes while half asleep. This little guy just sits on the desk and handles it all

pulmencr

56,202 Aufrufe • vor 2 Monaten

you’re sitting on a goldmine of content - it’s just trapped inside your sales calls this AI agent pulls the transcript and spits out: - a blog post - a LinkedIn post - a Twitter thread - an Instagram caption all in your tone. all platform-native. all actually good. just drop in the call and get 4 ready-to-post pieces in under 60 seconds built the whole thing in n8n after realizing I kept saying great stuff on calls… but never using it want the full workflow? comment “CALL,” follow, + repost and I’ll send it over (must be following so I can DM)

you’re sitting on a goldmine of content - it’s just trapped inside your sales calls this AI agent pulls the transcript and spits out: - a blog post - a LinkedIn post - a Twitter thread - an Instagram caption all in your tone. all platform-native. all actually good. just drop in the call and get 4 ready-to-post pieces in under 60 seconds built the whole thing in n8n after realizing I kept saying great stuff on calls… but never using it want the full workflow? comment “CALL,” follow, + repost and I’ll send it over (must be following so I can DM)

J.B.

22,537 Aufrufe • vor 1 Jahr

A friend of mine works in market research Her job involves calling different brands, asking questions, and collecting insights that later turn into reports. She needs to capture everything accurately while the conversation is still happening Earlier, she used to note everything down manually during the call. It was stressful and she often had to spend extra time later cleaning up messy notes. Now she does something much simpler. She keeps the call on speaker on her laptop, turns on the Typeless mic, and it converts the conversation into clean notes in real time. After the call, she just asks it to turn those notes into a proper summary or email for her team. No scrambling through rough notes. No spending another 20 minutes rewriting everything. She even does the same thing on mobile when she needs to send a quick update after a call. What makes it powerful is how it handles real speech. We don’t talk in perfect sentences. We pause, repeat ourselves, change direction mid-sentence, and add filler words. Typeless understands that. It removes filler words, fixes repetitions, catches mid-sentence changes, and turns messy speech into clean, ready to send text, whether that’s an email, a bullet list, notes, or a message. One feature that really stands out is voice based editing. You can refine sentences, change the tone, or tweak wording just by speaking again. No typing needed. It works anywhere you write WhatsApp, Slack, email, Notes, even ChatGPT and supports 100+ languages with a personal dictionary. Privacy is also a big focus. The platform is HIPAA compliant and GDPR compliant, with zero cloud data retention. Your data is never used to train AI models, and your history stays stored locally on your device. For professionals working with sensitive information, that matters a lot. They’re also working toward SOC 2 Type II and ISO 27001 compliance. Speaking your thoughts and getting polished writing instantly is honestly much faster than typing everything. That’s basically what Typeless does. Try it here: Android: (check the video 👇) Trust center:

A friend of mine works in market research Her job involves calling different brands, asking questions, and collecting insights that later turn into reports. She needs to capture everything accurately while the conversation is still happening Earlier, she used to note everything down manually during the call. It was stressful and she often had to spend extra time later cleaning up messy notes. Now she does something much simpler. She keeps the call on speaker on her laptop, turns on the Typeless mic, and it converts the conversation into clean notes in real time. After the call, she just asks it to turn those notes into a proper summary or email for her team. No scrambling through rough notes. No spending another 20 minutes rewriting everything. She even does the same thing on mobile when she needs to send a quick update after a call. What makes it powerful is how it handles real speech. We don’t talk in perfect sentences. We pause, repeat ourselves, change direction mid-sentence, and add filler words. Typeless understands that. It removes filler words, fixes repetitions, catches mid-sentence changes, and turns messy speech into clean, ready to send text, whether that’s an email, a bullet list, notes, or a message. One feature that really stands out is voice based editing. You can refine sentences, change the tone, or tweak wording just by speaking again. No typing needed. It works anywhere you write WhatsApp, Slack, email, Notes, even ChatGPT and supports 100+ languages with a personal dictionary. Privacy is also a big focus. The platform is HIPAA compliant and GDPR compliant, with zero cloud data retention. Your data is never used to train AI models, and your history stays stored locally on your device. For professionals working with sensitive information, that matters a lot. They’re also working toward SOC 2 Type II and ISO 27001 compliance. Speaking your thoughts and getting polished writing instantly is honestly much faster than typing everything. That’s basically what Typeless does. Try it here: Android: (check the video 👇) Trust center:

aditii

59,829 Aufrufe • vor 6 Monaten

A GUY MAKING $100K/MONTH WITH AI JUST SHOWED HIS ENTIRE SETUP. IT'S ONE FOLDER OF NOTES AND NOTHING ELSE no framework. no $500 course. he opens his screen and it's just obsidian - a plain notes app - wired into claude here's what he did: -> he pulled claude's memory files out of their default folder and dumped them into one vault -> had claude rename and merge them: 107 messy files collapsed into 17 clean ones -> every folder gets one master note that links to all the others that last part is the whole trick the agent reads the master note, follows the links and by the time it's done it has read every file in the folder. one instruction, full context here's the part most people miss: everyone's trying to make the AI smarter. he made the AI's memory smaller fewer files, better organized, all linked. the agent isn't scanning hundreds of notes anymore - it walks a path you built that's why his agent actually finishes jobs instead of forgetting what it was doing halfway through then he goes one step further: at the end of every session, the agent writes its own daily note. what it did, when, indexed at the top so it can find it again in seconds so he never re-explains anything. the agent looks up what it already did now he types "create a campaign for this offer" and walks away. it reads the product notes, reads the process notes, and comes back with the campaign done you don't need any of the complicated agent tools people are selling you. you need structure and instructions save this. the people winning with AI aren't using better models. they're just the only ones who bothered to organize what it remembers

A GUY MAKING $100K/MONTH WITH AI JUST SHOWED HIS ENTIRE SETUP. IT'S ONE FOLDER OF NOTES AND NOTHING ELSE no framework. no $500 course. he opens his screen and it's just obsidian - a plain notes app - wired into claude here's what he did: -> he pulled claude's memory files out of their default folder and dumped them into one vault -> had claude rename and merge them: 107 messy files collapsed into 17 clean ones -> every folder gets one master note that links to all the others that last part is the whole trick the agent reads the master note, follows the links and by the time it's done it has read every file in the folder. one instruction, full context here's the part most people miss: everyone's trying to make the AI smarter. he made the AI's memory smaller fewer files, better organized, all linked. the agent isn't scanning hundreds of notes anymore - it walks a path you built that's why his agent actually finishes jobs instead of forgetting what it was doing halfway through then he goes one step further: at the end of every session, the agent writes its own daily note. what it did, when, indexed at the top so it can find it again in seconds so he never re-explains anything. the agent looks up what it already did now he types "create a campaign for this offer" and walks away. it reads the product notes, reads the process notes, and comes back with the campaign done you don't need any of the complicated agent tools people are selling you. you need structure and instructions save this. the people winning with AI aren't using better models. they're just the only ones who bothered to organize what it remembers

Paone

23,833 Aufrufe • vor 2 Monaten

I just built my second brain with Hermes Agent. It remembers people, companies, projects, and decisions. It pulls from meeting notes, docs, anything I feed it, and cites the source every time. It's this repo called Gbrain, open-sourced by Y Combinator president Garry Tan. install it by pasting this into Hermes: "Install Gbrain following the instructions at Set up the embedding provider, then run gbrain init and gbrain doctor to verify." Add the MCP server so Hermes can call it directly: "Add the Gbrain MCP server and reload." The graph builds itself. Pattern matching on the people and companies already in your notes. 31 points more precise than plain vector search, by their own benchmark. The data from the video is made up so I could show you how it works. Hermes' memory remembers our conversations. Gbrain remembers my world. Open source. Runs local, no server, no Docker.

I just built my second brain with Hermes Agent. It remembers people, companies, projects, and decisions. It pulls from meeting notes, docs, anything I feed it, and cites the source every time. It's this repo called Gbrain, open-sourced by Y Combinator president Garry Tan. install it by pasting this into Hermes: "Install Gbrain following the instructions at Set up the embedding provider, then run gbrain init and gbrain doctor to verify." Add the MCP server so Hermes can call it directly: "Add the Gbrain MCP server and reload." The graph builds itself. Pattern matching on the people and companies already in your notes. 31 points more precise than plain vector search, by their own benchmark. The data from the video is made up so I could show you how it works. Hermes' memory remembers our conversations. Gbrain remembers my world. Open source. Runs local, no server, no Docker.

Alvaro Cintas

16,975 Aufrufe • vor 2 Monaten

We've just shipped private, automated meeting notes on your Mac, without a bot on the call. June detects meetings and takes notes for you. Turns it into decisions, action items, and follow ups. It all stays private. Stored locally. Open Source. MIT. Try it for free.

We've just shipped private, automated meeting notes on your Mac, without a bot on the call. June detects meetings and takes notes for you. Turns it into decisions, action items, and follow ups. It all stays private. Stored locally. Open Source. MIT. Try it for free.

OpenSoftware

334,907 Aufrufe • vor 3 Monaten

I built a simple internal chatbot that answers your team's SOPs and FAQs using your existing Google Drive docs in the walkthrough, I show: – how it finds the right doc for any question – how your team can “@” specific files (like in Slack) – how to set it up and share it with your team in minutes reply “chatbot” and I’ll send the full video + code (must be following)

I built a simple internal chatbot that answers your team's SOPs and FAQs using your existing Google Drive docs in the walkthrough, I show: – how it finds the right doc for any question – how your team can “@” specific files (like in Slack) – how to set it up and share it with your team in minutes reply “chatbot” and I’ll send the full video + code (must be following)

Tyler

27,282 Aufrufe • vor 1 Jahr

Larry Silverstein: "Pull it. And they made that decision to pull and we watched the building collapse." So when did they have time set up the explosives?

Larry Silverstein: "Pull it. And they made that decision to pull and we watched the building collapse." So when did they have time set up the explosives?

Green Lives Matter

16,994 Aufrufe • vor 2 Jahren

MICROSOFT JUST OPEN SOURCED THE ENTIRE AGENT RUNTIME, SO YOU NEVER HAVE TO BUILD ONE AGAIN a raw model only writes text. to make it a real agent it needs a loop, tools, memory, planning, approvals, and a way to know when it is done. Microsoft's Agent Harness is that whole layer, and it ships with everything on by default. you bring the model and the tools, it wires up the rest: > an automatic tool calling loop with iteration limits > history saved after every call, so a crash resumes instead of starting over > context compaction that summarizes when the window fills > a persistent todo list plus plan and execute modes > file memory that survives across turns > tool approval with standing rules and auto approval > OpenTelemetry tracing and web search built in > Agent Skills that load domain expertise on demand the mental model is simple: the model is the brain, the harness is the body around it. and it works with whatever you already run: > Azure OpenAI, OpenAI, Anthropic, and GitHub Copilot, and it hosts on Microsoft Foundry you spin one up in a single line: agent = create_harness_agent(client, name, instructions) open source, Python, .NET, and Go, stable since mid 2026. bookmark this. Microsoft already built the hard part of every AI agent, you just plug in the model.

MICROSOFT JUST OPEN SOURCED THE ENTIRE AGENT RUNTIME, SO YOU NEVER HAVE TO BUILD ONE AGAIN a raw model only writes text. to make it a real agent it needs a loop, tools, memory, planning, approvals, and a way to know when it is done. Microsoft's Agent Harness is that whole layer, and it ships with everything on by default. you bring the model and the tools, it wires up the rest: > an automatic tool calling loop with iteration limits > history saved after every call, so a crash resumes instead of starting over > context compaction that summarizes when the window fills > a persistent todo list plus plan and execute modes > file memory that survives across turns > tool approval with standing rules and auto approval > OpenTelemetry tracing and web search built in > Agent Skills that load domain expertise on demand the mental model is simple: the model is the brain, the harness is the body around it. and it works with whatever you already run: > Azure OpenAI, OpenAI, Anthropic, and GitHub Copilot, and it hosts on Microsoft Foundry you spin one up in a single line: agent = create_harness_agent(client, name, instructions) open source, Python, .NET, and Go, stable since mid 2026. bookmark this. Microsoft already built the hard part of every AI agent, you just plug in the model.

NO1ennn

69,163 Aufrufe • vor 1 Monat

A senior data scientist on call, 24/7. That's the dream. Well, these three built it. This is Data Scout, the agent that connects to Snowflake, Notion, Slack, Github, digs into your data, and sends recurring reports automatically. Honestly, one of our most-used agents at Notion. You can build one too.

A senior data scientist on call, 24/7. That's the dream. Well, these three built it. This is Data Scout, the agent that connects to Snowflake, Notion, Slack, Github, digs into your data, and sends recurring reports automatically. Honestly, one of our most-used agents at Notion. You can build one too.

Notion

40,481 Aufrufe • vor 5 Monaten

Mobile money agent left devastated after an unidentified man who allegedly came in to perform a transaction swapped bundles of GH¢200 notes totaling GH¢10,000 to bundles of GH¢1 cedi notes. According to reports, the customer interrupted a transaction, claiming they couldn't reach the recipient and asked the agent to stop. Moments after the agent refunded the cash, the customer allegedly returned with the money wrapped in the same black rubber. Believing it was the same cash she had already counted, the agent accepted it without conducting another verification and proceeded with the transaction. [🎥: MoMAG]

Mobile money agent left devastated after an unidentified man who allegedly came in to perform a transaction swapped bundles of GH¢200 notes totaling GH¢10,000 to bundles of GH¢1 cedi notes. According to reports, the customer interrupted a transaction, claiming they couldn't reach the recipient and asked the agent to stop. Moments after the agent refunded the cash, the customer allegedly returned with the money wrapped in the same black rubber. Believing it was the same cash she had already counted, the agent accepted it without conducting another verification and proceeded with the transaction. [🎥: MoMAG]

SIKAOFFICIAL🦍

389,206 Aufrufe • vor 2 Monaten

42 AI agents now run an entire sales department. Not one of them takes a salary. Hundreds of thousands in payroll. Gone in a weekend. At the top sits the chief sales officer. It delegates to every team’s manager. Outreach writes the emails, runs voice, fires SMS across 8 languages. RevOps wires HubSpot into the data warehouse. Research scores every lead and preps every call before a rep walks in. Then the part that breaks your brain. They talk to each other. The researcher hands context to the email writer. The call analyst feeds the follow up drafter. 42 specialists. 1 nervous system. This isn’t a chatbot. It’s an org chart that compounds while I sleep. And it’s day 1. Over the next 30 days this becomes a full company. Every department. Every role. Nobody on payroll. The last team I’ll ever have to hire. Want the full breakdown of all 42 agents and the org chart? Comment 42 and I’ll send it through.

42 AI agents now run an entire sales department. Not one of them takes a salary. Hundreds of thousands in payroll. Gone in a weekend. At the top sits the chief sales officer. It delegates to every team’s manager. Outreach writes the emails, runs voice, fires SMS across 8 languages. RevOps wires HubSpot into the data warehouse. Research scores every lead and preps every call before a rep walks in. Then the part that breaks your brain. They talk to each other. The researcher hands context to the email writer. The call analyst feeds the follow up drafter. 42 specialists. 1 nervous system. This isn’t a chatbot. It’s an org chart that compounds while I sleep. And it’s day 1. Over the next 30 days this becomes a full company. Every department. Every role. Nobody on payroll. The last team I’ll ever have to hire. Want the full breakdown of all 42 agents and the org chart? Comment 42 and I’ll send it through.

West Lord

29,206 Aufrufe • vor 3 Monaten