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.

just shipped taskmaster v0.15 🚀 → add-task is now context-aware 🔥 → research flag for parse-prd → new move command → analyze-complexity of specific tasks → add tasks without a PRD → ollama improved → celebrating 10k ⭐️ on GitHub +++ follow + bookmark + dive in 👀👇

Eyal Toledano

17,869 subscribers

15,474 views • 1 year 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

just shipped taskmaster v0.14 🚀 → know the cost of your taskmaster calls → ollama provider support → baseUrl support across roles → task complexity score in tasks → strong focus on fixes & stability → over 9,500⭐ on GitHub follow + bookmark + dive in 👀👇

just shipped taskmaster v0.14 🚀 → know the cost of your taskmaster calls → ollama provider support → baseUrl support across roles → task complexity score in tasks → strong focus on fixes & stability → over 9,500⭐ on GitHub follow + bookmark + dive in 👀👇

Eyal Toledano

14,360 views • 1 year ago

just shipped taskmaster v0.16 🚀 → .taskmaster folder → bedrock, @azure & Google AI vertex → TASK_MASTER_PROJECT_ROOT & BaseURL → semantic context for add-task → README <> task list sync + welcoming legend Jason Zhou to @taskmasterai follow + bookmark + dive in 👀👇

just shipped taskmaster v0.16 🚀 → .taskmaster folder → bedrock, @azure & Google AI vertex → TASK_MASTER_PROJECT_ROOT & BaseURL → semantic context for add-task → README <> task list sync + welcoming legend Jason Zhou to @taskmasterai follow + bookmark + dive in 👀👇

Eyal Toledano

69,502 views • 1 year ago

just shipped taskmaster v0.17 🚀 → multi-context task management via tagged lists → `research` MCP tool with interactive CLI → get-task/show upgrades & batch ops → --append for update-task → save research to docs → multi-status filters follow + bookmark + dive in 👀👇

just shipped taskmaster v0.17 🚀 → multi-context task management via tagged lists → `research` MCP tool with interactive CLI → get-task/show upgrades & batch ops → --append for update-task → save research to docs → multi-status filters follow + bookmark + dive in 👀👇

Eyal Toledano

37,012 views • 1 year ago

we've shipped taskmaster v0.27 🚀 → introducing `start` command → taskmaster auto-installs updates → grok CLI support GrokCli → rules & config Kilo → smart parent task status → move tasks across tags → gpt-oss support + loads more follow + retweet + bookmark 👀👇

we've shipped taskmaster v0.27 🚀 → introducing `start` command → taskmaster auto-installs updates → grok CLI support GrokCli → rules & config Kilo → smart parent task status → move tasks across tags → gpt-oss support + loads more follow + retweet + bookmark 👀👇

Eyal Toledano

11,376 views • 10 months ago

just shipped taskmaster v0.18 🚀 → Claude Code Community provider without API keys → init with IDE-specific profiles for Cursor Claude Code Community @windsurf Cline Visual Studio Code TRAE → cleaner git repos & Python support → PROVIDER_BASE_URL follow + bookmark + dive in 👀👇

just shipped taskmaster v0.18 🚀 → Claude Code Community provider without API keys → init with IDE-specific profiles for Cursor Claude Code Community @windsurf Cline Visual Studio Code TRAE → cleaner git repos & Python support → PROVIDER_BASE_URL follow + bookmark + dive in 👀👇

Eyal Toledano

85,994 views • 1 year ago

Are you ready? AGIX Ecosystem has Arrived!🚀 [1/10] AGIX is the world’s first AGI & AI Agents marketplace in Web3. Unlike ChatGPT, which handles single tasks, AGIX powers multi-agent systems to tackle complex tasks — all orchestrated by our AGI core. 🔍 Watch Our CTO Vlad Demonstrate the Power of AGIX: • Text-to-Action — See AGIX not just analyze, but act in real-time. • Multi-Tool Integration — Hundreds of tools at its command for any task. • Multi-Cycle Task Processing — Self-refining AI that iterates for perfection. • Full Autonomy — Give AGIX a complex task and let it handle the rest, with live updates in Telegram. It's the beginning of an AI Agents Economy.

Are you ready? AGIX Ecosystem has Arrived!🚀 [1/10] AGIX is the world’s first AGI & AI Agents marketplace in Web3. Unlike ChatGPT, which handles single tasks, AGIX powers multi-agent systems to tackle complex tasks — all orchestrated by our AGI core. 🔍 Watch Our CTO Vlad Demonstrate the Power of AGIX: • Text-to-Action — See AGIX not just analyze, but act in real-time. • Multi-Tool Integration — Hundreds of tools at its command for any task. • Multi-Cycle Task Processing — Self-refining AI that iterates for perfection. • Full Autonomy — Give AGIX a complex task and let it handle the rest, with live updates in Telegram. It's the beginning of an AI Agents Economy.

AGIX | $AGX

95,347 views • 1 year ago

Super excited to share 🧠MLGym 🦾 – the first Gym environment for AI Research Agents 🤖🔬 We introduce MLGym and MLGym-Bench, a new framework and benchmark for evaluating and developing LLM agents on AI research tasks. The key contributions of our work are: 🕹️ Enables the exploration of different training algorithms for AI Research Agents such as RL 🛠️ Provides a flexible evaluation framework that can accommodate different artifacts such as models, algorithms, or predictions 🤖 Allows researchers to evaluate any model without the need to develop a custom agentic harness 🎯 Introduces 13 diverse open-ended AI Research tasks for evaluating AI Research Agents on a wide range of domains such as computer vision, natural language processing, reinforcement learning, game theory, and logical reasoning. 📈 Proposes a new evaluation metric for AI Research Agents MLGym makes it easy to: 1) Add new tasks 2) Evaluate new models 3) Integrate new agents Check out a video of the MLGym Agent to see how it performs the full pipeline of idea generation💡, implementation 👩‍💻, experimentation 👩‍🔬, and iteration 🔄 to improve on ML tasks. Huge thanks to the exceptionally talented Deepak Nathani who led this work and to all the other amazing collaborators who made this possible 🙏🫶🚀

Super excited to share 🧠MLGym 🦾 – the first Gym environment for AI Research Agents 🤖🔬 We introduce MLGym and MLGym-Bench, a new framework and benchmark for evaluating and developing LLM agents on AI research tasks. The key contributions of our work are: 🕹️ Enables the exploration of different training algorithms for AI Research Agents such as RL 🛠️ Provides a flexible evaluation framework that can accommodate different artifacts such as models, algorithms, or predictions 🤖 Allows researchers to evaluate any model without the need to develop a custom agentic harness 🎯 Introduces 13 diverse open-ended AI Research tasks for evaluating AI Research Agents on a wide range of domains such as computer vision, natural language processing, reinforcement learning, game theory, and logical reasoning. 📈 Proposes a new evaluation metric for AI Research Agents MLGym makes it easy to: 1) Add new tasks 2) Evaluate new models 3) Integrate new agents Check out a video of the MLGym Agent to see how it performs the full pipeline of idea generation💡, implementation 👩‍💻, experimentation 👩‍🔬, and iteration 🔄 to improve on ML tasks. Huge thanks to the exceptionally talented Deepak Nathani who led this work and to all the other amazing collaborators who made this possible 🙏🫶🚀

Roberta Raileanu

105,052 views • 1 year ago

🚀 Introducing Queue: A Smarter Way to Build with Replit Agent Today, we’re excited to launch Queue, a powerful new capability designed to enhance your experience with Replit Agent. With Queue, you can submit multiple tasks while your Agent is still working—no more waiting around or breaking your flow. As each task completes, the Agent automatically moves to the next. You can: ✅ Add multiple tasks for the Agent to process in order. ✅ Edit, reorder, or reprioritize messages on the fly. ✅ Attach files and adjust settings—including toggling extended thinking or selecting a different model—per task.

🚀 Introducing Queue: A Smarter Way to Build with Replit Agent Today, we’re excited to launch Queue, a powerful new capability designed to enhance your experience with Replit Agent. With Queue, you can submit multiple tasks while your Agent is still working—no more waiting around or breaking your flow. As each task completes, the Agent automatically moves to the next. You can: ✅ Add multiple tasks for the Agent to process in order. ✅ Edit, reorder, or reprioritize messages on the fly. ✅ Attach files and adjust settings—including toggling extended thinking or selecting a different model—per task.

Replit ⠕

75,576 views • 1 year ago

update: i gave chatgpt full access to the taskmaster github repo and asked it to produce a thorough investigation of the product, the code and architecture, and to pay special attention to LLM-powered test generation, which I particularly care about right now it responded with an 8-page Mackenzie consultant-level report that dives deep into everything I asked and more one really solid thing it did that taskmaster makes possible is I asked it to go through all of my Taskmaster task files and to give recommendations on which are most high value/high priority. While I already had researched these tasks, it does add value to get another perspective i’ve been used to this with o1 pro, but to see the level of reasoning achieved against a growing and reasonably large code base blew my mind overall, it provided: - general audit and best practice recommendations - suggestions for new features - recommendations for how to consolidate MCP tools from 25 to 5-6 - review of open npm i task-master-ai tasks - exploration of LLM powered test generation with full references of where it grabbed that information absolutely incredible

update: i gave chatgpt full access to the taskmaster github repo and asked it to produce a thorough investigation of the product, the code and architecture, and to pay special attention to LLM-powered test generation, which I particularly care about right now it responded with an 8-page Mackenzie consultant-level report that dives deep into everything I asked and more one really solid thing it did that taskmaster makes possible is I asked it to go through all of my Taskmaster task files and to give recommendations on which are most high value/high priority. While I already had researched these tasks, it does add value to get another perspective i’ve been used to this with o1 pro, but to see the level of reasoning achieved against a growing and reasonably large code base blew my mind overall, it provided: - general audit and best practice recommendations - suggestions for new features - recommendations for how to consolidate MCP tools from 25 to 5-6 - review of open npm i task-master-ai tasks - exploration of LLM powered test generation with full references of where it grabbed that information absolutely incredible

Eyal Toledano

12,672 views • 1 year ago

Tracking Anything with Decoupled Video Segmentation paper page: Training data for video segmentation are expensive to annotate. This impedes extensions of end-to-end algorithms to new video segmentation tasks, especially in large-vocabulary settings. To 'track anything' without training on video data for every individual task, we develop a decoupled video segmentation approach (DEVA), composed of task-specific image-level segmentation and class/task-agnostic bi-directional temporal propagation. Due to this design, we only need an image-level model for the target task (which is cheaper to train) and a universal temporal propagation model which is trained once and generalizes across tasks. To effectively combine these two modules, we use bi-directional propagation for (semi-)online fusion of segmentation hypotheses from different frames to generate a coherent segmentation. We show that this decoupled formulation compares favorably to end-to-end approaches in several data-scarce tasks including large-vocabulary video panoptic segmentation, open-world video segmentation, referring video segmentation, and unsupervised video object segmentation.

Tracking Anything with Decoupled Video Segmentation paper page: Training data for video segmentation are expensive to annotate. This impedes extensions of end-to-end algorithms to new video segmentation tasks, especially in large-vocabulary settings. To 'track anything' without training on video data for every individual task, we develop a decoupled video segmentation approach (DEVA), composed of task-specific image-level segmentation and class/task-agnostic bi-directional temporal propagation. Due to this design, we only need an image-level model for the target task (which is cheaper to train) and a universal temporal propagation model which is trained once and generalizes across tasks. To effectively combine these two modules, we use bi-directional propagation for (semi-)online fusion of segmentation hypotheses from different frames to generate a coherent segmentation. We show that this decoupled formulation compares favorably to end-to-end approaches in several data-scarce tasks including large-vocabulary video panoptic segmentation, open-world video segmentation, referring video segmentation, and unsupervised video object segmentation.

AK

305,667 views • 2 years ago

The robot is learning several novel tasks instantly, after just ONE demonstration each... Instant Policy makes it possible: no extra training, no weight updates, just pure in-context learning. It just got accepted at ICLR 2025, and it’s changing how robots learn. With just a single demo, a robot can pick up a new task and start performing it right away. Why this is a big deal: ✅ Learns tasks instantly with just one or a few demonstrations ✅ Improves over time as more demonstrations are given ✅ Uses simulation-based training with “pseudo-demonstrations” for scalability ✅ Can transfer skills across different robots and even follow language-defined tasks It brings in-context learning to robotics, opening up new possibilities for flexible, real-world automation. You can try it yourself: code and weights are available at • • • • Thank you to Edward Johns, Director of the Robot Learning Lab at Imperial College for sharing their work! 🙏

The robot is learning several novel tasks instantly, after just ONE demonstration each... Instant Policy makes it possible: no extra training, no weight updates, just pure in-context learning. It just got accepted at ICLR 2025, and it’s changing how robots learn. With just a single demo, a robot can pick up a new task and start performing it right away. Why this is a big deal: ✅ Learns tasks instantly with just one or a few demonstrations ✅ Improves over time as more demonstrations are given ✅ Uses simulation-based training with “pseudo-demonstrations” for scalability ✅ Can transfer skills across different robots and even follow language-defined tasks It brings in-context learning to robotics, opening up new possibilities for flexible, real-world automation. You can try it yourself: code and weights are available at • • • • Thank you to Edward Johns, Director of the Robot Learning Lab at Imperial College for sharing their work! 🙏

Ilir Aliu

46,285 views • 1 year ago

🚀 We just shipped a big upgrade to our imitation-learning-in-simulation playground in LeRobot, built together with Lightwheel ! You can now teleoperate robots in sim (keyboard or real robot) and collect training demos instantly. This makes it possible to run real IL research on harder, more realistic manipulation tasks, even if you don’t have hardware. New tasks: 🟠 pick orange to the plate 🧺 fold cloth (yes!) 📦 pick 2 “e” toys to the box 🔴 lift red cube Through our partnership with Lightwheel, LeIsaac was integrated into EnvHub on day one, a best-practice integration that strengthens both ecosystems and pushes simulation-first robotics forward. 👉 Load a task via EnvHub, start teleop-ing, record your demo, upload the data on the hub and you’re ready to train or test.

🚀 We just shipped a big upgrade to our imitation-learning-in-simulation playground in LeRobot, built together with Lightwheel ! You can now teleoperate robots in sim (keyboard or real robot) and collect training demos instantly. This makes it possible to run real IL research on harder, more realistic manipulation tasks, even if you don’t have hardware. New tasks: 🟠 pick orange to the plate 🧺 fold cloth (yes!) 📦 pick 2 “e” toys to the box 🔴 lift red cube Through our partnership with Lightwheel, LeIsaac was integrated into EnvHub on day one, a best-practice integration that strengthens both ecosystems and pushes simulation-first robotics forward. 👉 Load a task via EnvHub, start teleop-ing, record your demo, upload the data on the hub and you’re ready to train or test.

LeRobot

44,647 views • 8 months ago

540,000 lines of Rails is a brutal way to learn the agent era Garry's List shipped with: > 262,000 lines of app code > 276,000 lines of tests > 127 background jobs > 33 cron jobs > a 1,778-line claim verifier Then GStack hit 105k GitHub stars Do not copy the Rails pattern 1. Finish the task with an agent 2. Say `skillify it` 3. Save the markdown skill 4. Keep TypeScript only for deterministic I/O 5. Add a unit test for the code 6. Add an LLM eval for the skill 7. Add an integration test across both 8. Add a resolver so the agent loads the skill next time A Rails feature serves one app A skill pack can run against the next hackathon, repo, customer inbox, research folder, or QA run Audit your repo this week Cut code that only exists to: > re-check model output > retry the same model call > validate what the model already has context for > schedule work because the agent has no reusable procedure > preserve a one-off prompt in chat history Keep the code that touches files, APIs, auth, billing, devices, databases Move the judgment into markdown Test the markdown Reuse the skill

540,000 lines of Rails is a brutal way to learn the agent era Garry's List shipped with: > 262,000 lines of app code > 276,000 lines of tests > 127 background jobs > 33 cron jobs > a 1,778-line claim verifier Then GStack hit 105k GitHub stars Do not copy the Rails pattern 1. Finish the task with an agent 2. Say `skillify it` 3. Save the markdown skill 4. Keep TypeScript only for deterministic I/O 5. Add a unit test for the code 6. Add an LLM eval for the skill 7. Add an integration test across both 8. Add a resolver so the agent loads the skill next time A Rails feature serves one app A skill pack can run against the next hackathon, repo, customer inbox, research folder, or QA run Audit your repo this week Cut code that only exists to: > re-check model output > retry the same model call > validate what the model already has context for > schedule work because the agent has no reusable procedure > preserve a one-off prompt in chat history Keep the code that touches files, APIs, auth, billing, devices, databases Move the judgment into markdown Test the markdown Reuse the skill

Alpha Batcher

35,693 views • 2 months ago

HERMES AGENT JUST GOT A RESEARCH BRAIN. connect it to NotebookLM and it becomes a 247 research system that builds knowledge over time. it resources. connects ideas across sources. uses that context in every real task. setup in 4 steps: 1. install Hermes with MCP enabled 2. download NotebookLM skill from GitHub 3. add endpoint to MCP servers in config 4. restart Hermes what your agent can now do: → query your notebooks instead of the open web → connect ideas across multiple sources → build knowledge session after session → use your curated context in every /goal → zero hallucinations from your own verified sources ask it anything — it answers from what YOU already know. full setup guide — MCP, models, use cases — in the article 👇

HERMES AGENT JUST GOT A RESEARCH BRAIN. connect it to NotebookLM and it becomes a 247 research system that builds knowledge over time. it resources. connects ideas across sources. uses that context in every real task. setup in 4 steps: 1. install Hermes with MCP enabled 2. download NotebookLM skill from GitHub 3. add endpoint to MCP servers in config 4. restart Hermes what your agent can now do: → query your notebooks instead of the open web → connect ideas across multiple sources → build knowledge session after session → use your curated context in every /goal → zero hallucinations from your own verified sources ask it anything — it answers from what YOU already know. full setup guide — MCP, models, use cases — in the article 👇

YanXbt

186,667 views • 2 months ago

Most motion papers tailor one controller to one specific task. This year at SIGGRAPH, our research team asks: can motor control itself be pretrained and reused? Generative Pretrained Controllers, or GPC, turn motor skills into a vocabulary of discrete tokens and train a transformer-based generative controller through next-token prediction. Just like GPT, the same pretrained controller can then be fine-tuned to solve new tasks. Trained on 600+ hours of motion, GPC runs in real-time inside a physics simulation, producing natural and physically grounded behaviors for interactive control.

Most motion papers tailor one controller to one specific task. This year at SIGGRAPH, our research team asks: can motor control itself be pretrained and reused? Generative Pretrained Controllers, or GPC, turn motor skills into a vocabulary of discrete tokens and train a transformer-based generative controller through next-token prediction. Just like GPT, the same pretrained controller can then be fine-tuned to solve new tasks. Trained on 600+ hours of motion, GPC runs in real-time inside a physics simulation, producing natural and physically grounded behaviors for interactive control.

NVIDIA AI

175,153 views • 1 month ago

Share Claude Code sessions with this command 👇 Just shipped v1.26.1 of claude-code-templates because I kept needing to hand off work mid-session. Run: npx claude-code-templates@latest --chats - All your sessions grouped by chat - Download any conversation To share a session: - Share the file, and add it to your project - Then run: claude "read @ chat-file .md and continue" Or if you are in the same account click on "Resume" and you will see the command: claude --resume session_id Built this to keep context flowing across the team without losing momentum.

Share Claude Code sessions with this command 👇 Just shipped v1.26.1 of claude-code-templates because I kept needing to hand off work mid-session. Run: npx claude-code-templates@latest --chats - All your sessions grouped by chat - Download any conversation To share a session: - Share the file, and add it to your project - Then run: claude "read @ chat-file .md and continue" Or if you are in the same account click on "Resume" and you will see the command: claude --resume session_id Built this to keep context flowing across the team without losing momentum.

Daniel San

10,787 views • 9 months ago

Grok just got a massive productivity upgrade xAI officially launched Connectors on Grok today Instead of copy-pasting between tabs, you can now link your apps directly to Grok for seamless, end-to-end workflows Just sign in to Grok web, click the + button, select Connectors, and click + Add connector to get started You can now connect: → Microsoft 365, SharePoint & OneDrive (read/edit files & docs) → Google Workspace & Outlook (draft emails, update calendars) → GitHub (search code, summarize PRs) → Notion & Linear (edit wikis, manage tasks) → Bring Your Own MCP (plug in custom internal tools) Live now on the web, and launching on Grok iOS and Android apps very soon

Grok just got a massive productivity upgrade xAI officially launched Connectors on Grok today Instead of copy-pasting between tabs, you can now link your apps directly to Grok for seamless, end-to-end workflows Just sign in to Grok web, click the + button, select Connectors, and click + Add connector to get started You can now connect: → Microsoft 365, SharePoint & OneDrive (read/edit files & docs) → Google Workspace & Outlook (draft emails, update calendars) → GitHub (search code, summarize PRs) → Notion & Linear (edit wikis, manage tasks) → Bring Your Own MCP (plug in custom internal tools) Live now on the web, and launching on Grok iOS and Android apps very soon

X Freeze

160,096 views • 2 months ago

SonicScribe V1 is LIVE! We are excited to announce a true breakthrough in confidential audio data handling – SonicScribe is now available for all users! This revolutionary technology transforms how organizations work with sensitive audio content by: • Accurate Transcriptions – with automatic redaction of personal identifiers. • Smart Q&A – ask natural language questions and receive context-aware, privacy-preserving answers. • Providing answers – that preserve context while protecting privacy. • Safe Search – analyze content without ever exposing confidential details. For the first time, organizations handling confidential information can safely allow employees to perform critical tasks without privacy risks. Law firms can analyze case recordings, HR departments can process interview transcripts, and customer service teams can mine call data – all while sensitive details remain protected. This technology bridges a critical gap that has long existed in regulated industries. Now you can extract maximum value from confidential audio while maintaining complete compliance with privacy regulations. 👉 Try it now:

SonicScribe V1 is LIVE! We are excited to announce a true breakthrough in confidential audio data handling – SonicScribe is now available for all users! This revolutionary technology transforms how organizations work with sensitive audio content by: • Accurate Transcriptions – with automatic redaction of personal identifiers. • Smart Q&A – ask natural language questions and receive context-aware, privacy-preserving answers. • Providing answers – that preserve context while protecting privacy. • Safe Search – analyze content without ever exposing confidential details. For the first time, organizations handling confidential information can safely allow employees to perform critical tasks without privacy risks. Law firms can analyze case recordings, HR departments can process interview transcripts, and customer service teams can mine call data – all while sensitive details remain protected. This technology bridges a critical gap that has long existed in regulated industries. Now you can extract maximum value from confidential audio while maintaining complete compliance with privacy regulations. 👉 Try it now:

ZKCrypt AI

17,101 views • 1 year ago

🌟 Simple guide to SimpleChain Testnet 🪂🪂 💰 Cost: $0 💵 Funding: $15M 🪙 Rewards: Confirmed Simple Chain is an L1 chain that requires its own token as gas for the mainnet launch, so don't forget to give it a try Steps: 🔗 ✅ Add extension ✅ Create new wallet ✅ Open wallet ✅ Copy SRW address 🔗 ✅ Paste the copied address ✅ Claim faucet ✅ Try later if it's not going through 🔗 ✅ Connect wallet ✅ Go to profile ✅ Connect X, Discord and TG ✅ Complete daily check-in ✅ Complete available tasks ✅ Done Don't forget to complete available tasks and daily check-ins That's all for now and will update later once new tasks appear Follow and turn on notifications for updates 🤍 Like 🔁 Retweet

CCT

17,428 views • 3 months ago

$ANTHROPIC JUST QUIETLY SHIPPED A FEATURE THAT LETS CLAUDE SPAWN A WHOLE TEAM OF AGENTS THAT MESSAGE EACH OTHER AND REVIEW EACH OTHER'S WORK. It's a Claude Code feature called agent teams. The team lead spawns multiple agents that share a task list and message each other directly, not subagents reporting back, actual peers. In the demo a QA agent caught three bugs, sent the work back to the front-end and back-end devs, they fixed it, app shipped in one pass. How to run it: 1. Enable it. Needs Claude Code v2.1.32+. Add to settings.json: "env": { "CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1" }. Or paste that to Claude and say "add this to my settings." Restart. 2. Prompt in plain English. Start with a goal (agents wake with zero context), then "create a team of 3 using Sonnet," describe each role, its deliverable, and who it messages when done. 3. The rules: each agent owns its own files, define exact outputs, name who talks to who, keep it to 3-5 agents. Use it for complex work with separate parts running in parallel. Skip it for simple or sequential tasks, teams cost 3-4x the tokens. Bookmark this.$

ANTHROPIC JUST QUIETLY SHIPPED A FEATURE THAT LETS CLAUDE SPAWN A WHOLE TEAM OF AGENTS THAT MESSAGE EACH OTHER AND REVIEW EACH OTHER'S WORK. It's a Claude Code feature called agent teams. The team lead spawns multiple agents that share a task list and message each other directly, not subagents reporting back, actual peers. In the demo a QA agent caught three bugs, sent the work back to the front-end and back-end devs, they fixed it, app shipped in one pass. How to run it: 1. Enable it. Needs Claude Code v2.1.32+. Add to settings.json: "env": { "CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1" }. Or paste that to Claude and say "add this to my settings." Restart. 2. Prompt in plain English. Start with a goal (agents wake with zero context), then "create a team of 3 using Sonnet," describe each role, its deliverable, and who it messages when done. 3. The rules: each agent owns its own files, define exact outputs, name who talks to who, keep it to 3-5 agents. Use it for complex work with separate parts running in parallel. Skip it for simple or sequential tasks, teams cost 3-4x the tokens. Bookmark this.

Yarchi

460,553 views • 1 month ago