Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Introducing Branch Split any conversation into a parallel session that inherits your full context, while the original stays intact. A single research session can become a report, a slide deck, and an investor brief. Now you just branch. Same context, new direction.

Manus

248,783 subscribers

58,098 views • 2 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

You can now resume your terminal sessions in the Claude Code desktop app. Type /resume to pick any session you started from the CLI. The session continues in the app with the full conversation and context intact.

You can now resume your terminal sessions in the Claude Code desktop app. Type /resume to pick any session you started from the CLI. The session continues in the app with the full conversation and context intact.

ClaudeDevs

648,455 views • 19 days ago

Introducing side chats, a new way to ask questions and explore ideas without interrupting your main conversation. Each side chat is a durable agent conversation you can @-mention to bring context back into the main thread.

Sensitive content

Introducing side chats, a new way to ask questions and explore ideas without interrupting your main conversation. Each side chat is a durable agent conversation you can @-mention to bring context back into the main thread.

Cursor

435,147 views • 2 months ago

📡Web Scraping is Live in the app and API You can now turn any URL into AI context on Venice Just include a URL in your prompt, and Venice will automatically scrape the page to include as context for your request

📡Web Scraping is Live in the app and API You can now turn any URL into AI context on Venice Just include a URL in your prompt, and Venice will automatically scrape the page to include as context for your request

Venice

39,679 views • 10 months ago

Introducing cmux Vault: cmux now has a right sidebar with a vault pane where you can see every single Codex, Claude Code, OpenCode, and Pi session You can search through all sessions with full-text search, and drag them directly into your workspace Out in v0.64+

Introducing cmux Vault: cmux now has a right sidebar with a vault pane where you can see every single Codex, Claude Code, OpenCode, and Pi session You can search through all sessions with full-text search, and drag them directly into your workspace Out in v0.64+

Lawrence Chen

32,394 views • 4 months ago

We took a 30B model and split it in two to write tokens in parallel instead of one at a time. Introducing Nemotron-Labs-TwoTower: a diffusion language model from NVIDIA Research adapted from Nemotron-3-Nano-30B-A3B. Here’s how it works: one half holds the context, the other writes the tokens, with both reusing the pretrained model instead of training a new one from scratch. We found it kept 98.7% of the original model’s quality at 2.42× faster generation.

We took a 30B model and split it in two to write tokens in parallel instead of one at a time. Introducing Nemotron-Labs-TwoTower: a diffusion language model from NVIDIA Research adapted from Nemotron-3-Nano-30B-A3B. Here’s how it works: one half holds the context, the other writes the tokens, with both reusing the pretrained model instead of training a new one from scratch. We found it kept 98.7% of the original model’s quality at 2.42× faster generation.

NVIDIA AI

762,173 views • 2 months ago

GEMMA 4 26B ON AN RTX 4060 WITH A 248K TOKEN CONTEXT WINDOW 20 tokens per second and a context window so large you can feed it entire codebases, books and research papers in a single prompt this is not a cloud api and not a server rack, this is a regular consumer gpu running locally with llama.cpp and q4_k_xl quantization 248k context on an 8gb vram card was not supposed to be possible and here it is just running on someone’s desk the article below covers exactly which tools and configs make this kind of setup work in 2026 ↓

GEMMA 4 26B ON AN RTX 4060 WITH A 248K TOKEN CONTEXT WINDOW 20 tokens per second and a context window so large you can feed it entire codebases, books and research papers in a single prompt this is not a cloud api and not a server rack, this is a regular consumer gpu running locally with llama.cpp and q4_k_xl quantization 248k context on an 8gb vram card was not supposed to be possible and here it is just running on someone’s desk the article below covers exactly which tools and configs make this kind of setup work in 2026 ↓

leopardracer

56,865 views • 3 months ago

INVESTOR DECKS ARE OFFICIALLY A ONE-PROMPT PROBLEM NOW goes from scattered brand notes and market research straight to an investor ready deck - eliminating manual slide layout, manual chart building, and manual re formatting between tools Here is the exact stack: Claude Design Canvas - reads a brand concept doc and a research outline, builds each slide type with real visual reasoning, not text dropped into a template Live Research Layer - for skills that support it, runs an actual web search per prompt, structures a narrative arc from current data instead of guessing at market size Design-System Ingestion - feed it a clean codebase or existing brand files, Claude infers your grid, type scale, and color system automatically instead of starting from a blank canvas Multi format Export - PPTX, PDF, HTML, or straight into Google Slides, so the deck opens correctly wherever the investor actually opens it Core advantage: set up a Project once with your colors, fonts, and style guide - every deck generated after that follows your brand automatically, no re-uploading brand assets slide by slide honest limit worth knowing: HTML to PPTX export is lossy text can flatten to images, fonts substitute, master slides drop If the deck's final home is PowerPoint, budget one manual pass to catch what didn't survive the conversion Pass a brand doc and a research outline Claude structures the narrative arc, builds every slide, and exports to whatever format the round actually needs zero blank canvas, zero manual chart rebuilding full below

YARD

32,780 views • 18 days ago

/autofix-pr now lets you kick off autofix straight from the command line. After finishing up a PR, just run /autofix-pr. It sends your session to the cloud so the PR autofixer has full context to address CI failures and comments. Hugely win for my productivity.

/autofix-pr now lets you kick off autofix straight from the command line. After finishing up a PR, just run /autofix-pr. It sends your session to the cloud so the PR autofixer has full context to address CI failures and comments. Hugely win for my productivity.

Noah Zweben

464,149 views • 5 months ago

Following up on dragging splits into other split locations and across windows, you can now also drag a split outside any valid windows to pull it out into a new window. It even works with undo/redo.

Following up on dragging splits into other split locations and across windows, you can now also drag a split outside any valid windows to pull it out into a new window. It even works with undo/redo.

Mitchell Hashimoto

35,754 views • 8 months ago

OK, so now we know. You can comfortably pack a swarm of 720 drones and their launch system into a single 20ft shipping container. For context Germany + France + UK have about 800 tanks between them. The world has changed.

OK, so now we know. You can comfortably pack a swarm of 720 drones and their launch system into a single 20ft shipping container. For context Germany + France + UK have about 800 tanks between them. The world has changed.

Object Zero

1,695,176 views • 11 months ago

Smol tip: /𝚏𝚘𝚛𝚔 is perfect for mid-task side quests. It spawns a subagent that inherits the whole conversation, runs in the background while you keep working on your main quest, and comes back with an update.

Smol tip: /𝚏𝚘𝚛𝚔 is perfect for mid-task side quests. It spawns a subagent that inherits the whole conversation, runs in the background while you keep working on your main quest, and comes back with an update.

Delba

112,303 views • 2 months ago

💡 Using Next.js's Intercepting Routes can be useful if you want to show a route while keeping the context for the current page! E.g. we can render an overlaying modal on soft navigation and a standalone page on hard navigation, which requires a full page reload:

💡 Using Next.js's Intercepting Routes can be useful if you want to show a route while keeping the context for the current page! E.g. we can render an overlaying modal on soft navigation and a standalone page on hard navigation, which requires a full page reload:

Lydia Hallie ✨

564,123 views • 3 years ago

THIS 38,000-STAR GITHUB REPO TURNS ONE AI AGENT INTO A REAL TEAM THAT CAN BRANCH, VERIFY ITS WORK AND WAIT FOR YOUR APPROVAL most people still run agents as one long chain where every step waits, one failure kills the run and the full workflow starts again Task → Planner → 5 Researchers in Parallel → Skeptic → Writer → Human Gate LangGraph gives every node one job while a shared state carries the findings, decisions and context through the entire system the skeptic can reject an unsupported finding and route the work back before it contaminates the final report, while independent branches keep moving if the run crashes, durable execution resumes from the saved state instead of rebuilding everything, then human-in-the-loop pauses the graph before anything expensive gets sent or published bookmark this repo and watch one prompt turn into an actual org chart for AI agents

THIS 38,000-STAR GITHUB REPO TURNS ONE AI AGENT INTO A REAL TEAM THAT CAN BRANCH, VERIFY ITS WORK AND WAIT FOR YOUR APPROVAL most people still run agents as one long chain where every step waits, one failure kills the run and the full workflow starts again Task → Planner → 5 Researchers in Parallel → Skeptic → Writer → Human Gate LangGraph gives every node one job while a shared state carries the findings, decisions and context through the entire system the skeptic can reject an unsupported finding and route the work back before it contaminates the final report, while independent branches keep moving if the run crashes, durable execution resumes from the saved state instead of rebuilding everything, then human-in-the-loop pauses the graph before anything expensive gets sent or published bookmark this repo and watch one prompt turn into an actual org chart for AI agents

Gipp 🦅

11,524 views • 1 month ago

Introducing agent-tail: a dev tool that makes your logs easily accessible to agents. npm i agent-tail Wrap dev server commands, add the Vite/Next plugin, install the skill, and your agent has full context of your browser, server, and any other logs. Link to full docs below ↓

Introducing agent-tail: a dev tool that makes your logs easily accessible to agents. npm i agent-tail Wrap dev server commands, add the Vite/Next plugin, install the skill, and your agent has full context of your browser, server, and any other logs. Link to full docs below ↓

Kyle Gill

29,414 views • 7 months ago

NEW in Dia: Inline Browsing Starting today, when you click links in a chat with Dia, the webpages will automatically render inline. No more context switching. No more tab mess. Now you can browse the web without breaking the flow of your conversation:

NEW in Dia: Inline Browsing Starting today, when you click links in a chat with Dia, the webpages will automatically render inline. No more context switching. No more tab mess. Now you can browse the web without breaking the flow of your conversation:

Dia

61,438 views • 1 year ago

There's a ton of stuff to do while starting a business, so we took inspiration from RTS games like Civ and built in a roadmap of an entire company. Kick off the next step while maintaining your full business context.

There's a ton of stuff to do while starting a business, so we took inspiration from RTS games like Civ and built in a roadmap of an entire company. Kick off the next step while maintaining your full business context.

andrew pignanelli

37,374 views • 4 months ago

Marjorie Taylor Greene and other conservatives are sharing a cut from a video that appears to show Joe Biden admitting to selling state secrets. Some conservatives are claiming that this shows Biden is corrupt while others are claiming that it is evidence of dementia. Context matters. They are deceiving you by only showing a small portion of the video which doesn’t show Biden saying “all kidding aside.” Context matters. This is irresponsible if you are taking sections of a video to sell an narrative you want people to believe. You can see the full context below.

Marjorie Taylor Greene and other conservatives are sharing a cut from a video that appears to show Joe Biden admitting to selling state secrets. Some conservatives are claiming that this shows Biden is corrupt while others are claiming that it is evidence of dementia. Context matters. They are deceiving you by only showing a small portion of the video which doesn’t show Biden saying “all kidding aside.” Context matters. This is irresponsible if you are taking sections of a video to sell an narrative you want people to believe. You can see the full context below.

Ed Krassenstein

3,489,690 views • 3 years ago

🚨 JUST IN: CHINA just released an AI EMPLOYEE that works 24X7 on its own. 100% OPEN SOURCE. It researches, codes, builds websites, creates slide decks, and generates videos. All by itself. All on your computer. It's called DeerFlow. You give it a task. It makes a plan, spins up its own team of sub-agents, and gets to work. You come back and there's a finished deliverable waiting. Not a draft. Not a summary. The actual thing. Not a chatbot. Not a research assistant. An AI with its own computer that works while you sleep. Here's what it does on its own: → Spawns multiple sub-agents in parallel, each tackling a different piece of your task, then combines everything into one finished output → Writes real code, runs it, reads the results, and fixes its own mistakes without asking you once → Builds slide decks, websites, full research reports, and data dashboards from scratch → Remembers you across sessions. Your writing style. Your tech stack. Your preferences. Gets better every time. → Reads files you upload, works with them inside its own filesystem, hands you clean finished outputs → Searches the web, runs commands, calls any tool you plug in Here's how it thinks: You give one instruction. The lead agent makes a plan. Sub-agents fan out and work in parallel. Results come back. Everything gets synthesized. You get a deliverable. A single research task might split into a dozen sub-agents, each exploring a different angle, then converge into one finished website with generated visuals. Here's the wildest part: DeerFlow 2.0 launched on February 28th 2026 and hit number 1 on all of GitHub Trending the same day. Version 2.0 was a complete rewrite. Zero shared code with version 1. Because users kept using it for things the team never intended. Data pipelines. Dashboards. Entire content workflows. The community told them what it needed to become. So they burned it down and rebuilt it. 22.7K GitHub stars. 2.7K forks. Built by ByteDance 100% Open Source. MIT License.

🚨 JUST IN: CHINA just released an AI EMPLOYEE that works 24X7 on its own. 100% OPEN SOURCE. It researches, codes, builds websites, creates slide decks, and generates videos. All by itself. All on your computer. It's called DeerFlow. You give it a task. It makes a plan, spins up its own team of sub-agents, and gets to work. You come back and there's a finished deliverable waiting. Not a draft. Not a summary. The actual thing. Not a chatbot. Not a research assistant. An AI with its own computer that works while you sleep. Here's what it does on its own: → Spawns multiple sub-agents in parallel, each tackling a different piece of your task, then combines everything into one finished output → Writes real code, runs it, reads the results, and fixes its own mistakes without asking you once → Builds slide decks, websites, full research reports, and data dashboards from scratch → Remembers you across sessions. Your writing style. Your tech stack. Your preferences. Gets better every time. → Reads files you upload, works with them inside its own filesystem, hands you clean finished outputs → Searches the web, runs commands, calls any tool you plug in Here's how it thinks: You give one instruction. The lead agent makes a plan. Sub-agents fan out and work in parallel. Results come back. Everything gets synthesized. You get a deliverable. A single research task might split into a dozen sub-agents, each exploring a different angle, then converge into one finished website with generated visuals. Here's the wildest part: DeerFlow 2.0 launched on February 28th 2026 and hit number 1 on all of GitHub Trending the same day. Version 2.0 was a complete rewrite. Zero shared code with version 1. Because users kept using it for things the team never intended. Data pipelines. Dashboards. Entire content workflows. The community told them what it needed to become. So they burned it down and rebuilt it. 22.7K GitHub stars. 2.7K forks. Built by ByteDance 100% Open Source. MIT License.

Kanika

738,832 views • 5 months ago

Claude just solved one of AI coding’s biggest problems. 🤯 It can now turn your entire codebase into: → an interactive architecture map for humans → a memory system for AI agents So the next coding agent instantly understands your APIs, components, auth, DB flows, and dependencies. No more dropping AI into a random repo with zero context. Your codebase can literally explain itself now.

Claude just solved one of AI coding’s biggest problems. 🤯 It can now turn your entire codebase into: → an interactive architecture map for humans → a memory system for AI agents So the next coding agent instantly understands your APIs, components, auth, DB flows, and dependencies. No more dropping AI into a random repo with zero context. Your codebase can literally explain itself now.

Radha Tripathi

23,136 views • 4 months ago

Bittensor is built for an agentic world. The new Bittensor documentation is machine-readable. Agents can load the full corpus into a context window, discover chain operations, preview them, and execute them. Builders also get a clearer path into the network, with a five-minute Quickstart, SDK and CLI guides, and migration resources. Full documentation in the first comment.

Bittensor is built for an agentic world. The new Bittensor documentation is machine-readable. Agents can load the full corpus into a context window, discover chain operations, preview them, and execute them. Builders also get a clearer path into the network, with a five-minute Quickstart, SDK and CLI guides, and migration resources. Full documentation in the first comment.

Openτensor Foundaτion

41,557 views • 1 month ago