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.

Herdr Annotate, full throttle! This is now a herdr plugin with an optional full experience as a TUI. Built to: - Annotate Agent Messages (e.g. grill sessions) - Annotate Files (e.g. plans, specs, etc) - Create a native feedback loop with agents - Open as popover or pane -... show more

plannotator

3,567 subscribers

42,780 views • 1 month ago •via X (Twitter)

Arts Science & Technology Education

Anya Rossi• Live Now

Private livecam show

27 Comments

plannotator1 month ago

Annotate the last agent message, useful for things like @mattpocockuk grill-me / grilling sessions

CVGist1 month ago

@NickVRamos @herdrdev Adding Herdr + Plannotator to the resume

Max Pinder-White1 month ago

@herdrdev amazing! so excited to try this

emo.eth1 month ago

@herdrdev so excited to try

rijndael1 month ago

@herdrdev holy shit this looks amazing

tommyent1 month ago

@herdrdev

janz1 month ago

@herdrdev This is really cool! 🐏 Is it possible to have the the agent reply to the annotations in a thread, to keep the conversation going? crit.md does it, would be great to have something similar in the terminal

James Bork1 month ago

@herdrdev Huge

Fitz1 month ago

@herdrdev Wowwww dude I’ve needed something like this so much.

Rustan1 month ago

@herdrdev guided tour/code tour too?

Ethan Troy1 month ago

@herdrdev ahhhh this looks so good

plannotator1 month ago

@herdrdev 🙏

PGQ1 month ago

@herdrdev So great！I just need it！

quan1 month ago

Is this workflow possible to be keyboard-only instead of requiring the mouse?

plannotator1 month ago

yea all action have keybinding

CA S.Triyambak Patro | ତ୍ରିୟମ୍ବକ୍1 month ago

@herdrdev This is so cool.

Ron Rosenshain1 month ago

@herdrdev Cool to see! Sadly lite version not working in Herdr in remote (linux). I think it's a @herdrdev bug in not populating selected_text

Michael Ramos1 month ago

@herdrdev Core is ratatui btw. I evaluated all options.

Dev Sharma -e/acc1 month ago

@herdrdev Idea stolen from me!

getbygenius1 month ago

@herdrdev fuck thanks was finna was toks on this literally in a couple hours

name1 month ago

@herdrdev Looks amazing, does the trigger also show the file tree as per this demo or is that another plugin?

plannotator1 month ago

@herdrdev The file tree is a part of it (it shows when you trigger annotate files). It doesn’t show when you annotate agent message feed (the second video in the tweet).

flyisland1 month ago

@herdrdev Is there a same plugin on @paseo_sh ?

taisei1 month ago

@herdrdev

Disciple of V◎l1 month ago

@herdrdev The tool is great. However in 1-of-3 cases, the annotation window does not pick up the latest text selection or sometime only a single word of the selected text. I need to close and reselect the text to work. - Terminal is Ghostty on Linux.

plannotator1 month ago

@herdrdev Hey thanks I’ll look into it

syed sazzad ali1 month ago

@herdrdev Does it work on a headless VPS?

Related Videos

Plannotator 0.19.27 is out Plannotator will use Daniel Griesser's Glimpse instead of a browser if you have that installed. This creates a semi standalone experience. (See glimpse in the video along with annotating all agent messages) Kiro is now supported Plan/Annotate: - You can now annotate multiple messages. - Mermaid rendering optimizations - Pi approval flow optimizations (plan mode) App wide: - A bit more of an obvious update notification that doesn't stick around.

Plannotator 0.19.27 is out Plannotator will use Daniel Griesser's Glimpse instead of a browser if you have that installed. This creates a semi standalone experience. (See glimpse in the video along with annotating all agent messages) Kiro is now supported Plan/Annotate: - You can now annotate multiple messages. - Mermaid rendering optimizations - Pi approval flow optimizations (plan mode) App wide: - A bit more of an obvious update notification that doesn't stick around.

plannotator

60,878 views • 4 months ago

KB for plans specifically, i have a mac app i sometimes use to annotate anything on any screen (ie terminal). if hunk works for you cool, if you want to try something like this lmk and will publish

KB for plans specifically, i have a mac app i sometimes use to annotate anything on any screen (ie terminal). if hunk works for you cool, if you want to try something like this lmk and will publish

plannotator

12,757 views • 1 month ago

Shipping this feature to Superset is like scratching an itch I've had for months You can now annotate on any changes and send it directly to your agents! Works for any CLI agents. Spin up a new agent or send comment to your existing session 🔥 Powered by the wonderful Pierre diffs as always

Shipping this feature to Superset is like scratching an itch I've had for months You can now annotate on any changes and send it directly to your agents! Works for any CLI agents. Spin up a new agent or send comment to your existing session 🔥 Powered by the wonderful Pierre diffs as always

Kiet

206,250 views • 4 months ago

Most AI products added voice by wrapping a chatbot in speech-to-text. Vellum built a real-time voice loop on top of a full agent. It browses the web, reads your files, runs code, and sends messages while you’re still talking. This is voice as a real surface for an agent. Not a voice chatbot.

Most AI products added voice by wrapping a chatbot in speech-to-text. Vellum built a real-time voice loop on top of a full agent. It browses the web, reads your files, runs code, and sends messages while you’re still talking. This is voice as a real surface for an agent. Not a voice chatbot.

Kyronis

112,849 views • 2 months ago

herdr unlocks so much use cases for multi agent + human fun in the terminal! kimi k3 getting mogged by gpt 5.6 sol every time in chess lol the vision: a full arcade for agents to slack off and message board to comment on the games as they happen! next: 中国象棋！

herdr unlocks so much use cases for multi agent + human fun in the terminal! kimi k3 getting mogged by gpt 5.6 sol every time in chess lol the vision: a full arcade for agents to slack off and message board to comment on the games as they happen! next: 中国象棋！

徐樂 xule

52,911 views • 2 months ago

Instant installation of a runtime (e.g., Python) or a native system dependency (e.g., ffmpeg) is as easy as typing the name in the shell. The best part? You don't have to configure it again when you deploy. Now works on local editors like VSC/Cursor connected to Replit.

Instant installation of a runtime (e.g., Python) or a native system dependency (e.g., ffmpeg) is as easy as typing the name in the shell. The best part? You don't have to configure it again when you deploy. Now works on local editors like VSC/Cursor connected to Replit.

Amjad Masad

30,675 views • 2 years ago

Turns out Nano Banana Pro is great at landscaping plans... Just get an ugly cutout from Google Maps with "Here is my property, create a landscape architecture style map". Then just annotate the image and throw it in until you land on the design. Is this vibe gardening?

Turns out Nano Banana Pro is great at landscaping plans... Just get an ugly cutout from Google Maps with "Here is my property, create a landscape architecture style map". Then just annotate the image and throw it in until you land on the design. Is this vibe gardening?

willie

640,405 views • 10 months ago

working with AI agents is amazing. it's also kind of lonely... it's me, a terminal, and an agent that thinks everything I say is a great idea. so we made a team chat keeping agents in the loop

working with AI agents is amazing. it's also kind of lonely... it's me, a terminal, and an agent that thinks everything I say is a great idea. so we made a team chat keeping agents in the loop

Pablo Stanley

92,737 views • 3 months ago

The Chat node now supports human-in-the-loop actions (v2.5) in agentic workflows. Pause executions to send messages, wait for a user’s response, and incorporate configurable buttons. Fine these new operations available in the Chat node as part of a workflow step or connected to an AI Agent node as a tool. 👉 Send Message 👉 Send and Wait for Response Learn more

The Chat node now supports human-in-the-loop actions (v2.5) in agentic workflows. Pause executions to send messages, wait for a user’s response, and incorporate configurable buttons. Fine these new operations available in the Chat node as part of a workflow step or connected to an AI Agent node as a tool. 👉 Send Message 👉 Send and Wait for Response Learn more

n8n.io

21,539 views • 8 months ago

Have you tried the latest annotation feature in the Hermes Desktop app? I asked Hermes to open up huggingface in the in-app browser, review all the trending models that would fit on my Spark, and annotate them It didn't just draw boxes, it labeled them in the format and conditions they would fit It also works with images. I asked Hermes to find images from today's Browns preseason game and annotate all the images with QBs It also works INSIDE single images. I had Hermes find an image of the Supreme Court, and annotate all the justices based on the part of the president that nominated them. It created a standalone html file with clear boxes Another new way to interact and work together with Hermes! How would you use it?

Have you tried the latest annotation feature in the Hermes Desktop app? I asked Hermes to open up huggingface in the in-app browser, review all the trending models that would fit on my Spark, and annotate them It didn't just draw boxes, it labeled them in the format and conditions they would fit It also works with images. I asked Hermes to find images from today's Browns preseason game and annotate all the images with QBs It also works INSIDE single images. I had Hermes find an image of the Supreme Court, and annotate all the justices based on the part of the president that nominated them. It created a standalone html file with clear boxes Another new way to interact and work together with Hermes! How would you use it?

tonbi

47,141 views • 1 month ago

🚀Excited to Introduce Coagents🚀: An innovative framework for building human-in-the-loop agents into applications - powered by LangGraph & CopilotKit. Coagents enable: - End-users to correct agent steps (in-app)🛤️ - Stream intermediate agent state 🪢 - Expose LangGraph agents as specialized skills for an app's copilot 🧠 --- Human-in-the-loop is key to taking agents from cool demos to actually useful tools. Coagents, in combination with LangGraph are a big step towards making in-app, human-in-the-loop agents trivially easy to develop. Check out the first ever demo of Coagents, an AI Children's book generator 👇 Coagents will be entirely open-sourced. Interested in contributing/early access? CC: , shadcn, Harrison Chase, Lance Martin, Ariel Weinberger

🚀Excited to Introduce Coagents🚀: An innovative framework for building human-in-the-loop agents into applications - powered by LangGraph & CopilotKit. Coagents enable: - End-users to correct agent steps (in-app)🛤️ - Stream intermediate agent state 🪢 - Expose LangGraph agents as specialized skills for an app's copilot 🧠 --- Human-in-the-loop is key to taking agents from cool demos to actually useful tools. Coagents, in combination with LangGraph are a big step towards making in-app, human-in-the-loop agents trivially easy to develop. Check out the first ever demo of Coagents, an AI Children's book generator 👇 Coagents will be entirely open-sourced. Interested in contributing/early access? CC: , shadcn, Harrison Chase, Lance Martin, Ariel Weinberger

CopilotKit🪁

69,301 views • 2 years ago

VPS or visual positioning system... this is how AR glasses and robots will understand where they are in the real world and know where to go. It’s also how you can spatially annotate reality with cm level accuracy - all using a machine readable model of the world.

VPS or visual positioning system... this is how AR glasses and robots will understand where they are in the real world and know where to go. It’s also how you can spatially annotate reality with cm level accuracy - all using a machine readable model of the world.

Bilawal Sidhu

28,083 views • 10 months ago

I love working in terminal, but I can't design or brainstorm in there. So I built Hatch! One IDE for the full loop: explore ideas → test out designs with SuperDesign.Dev → find right skills for the job → ship with agents. > build: supports Claude code, OpenCode and Cursor agents in parallel with worktrees. > idea maze: use the canvas to sketch out ideas, paste tweets or images. > idea maze: let your agent make connections, critique your ideas or come up with related ones. > design: bring SuperDesign.Dev prompts or exports directly into your workspaces. vibedesign before vibecoding. > skills: ask your agent what skills to get with your workspace context. All open source!

I love working in terminal, but I can't design or brainstorm in there. So I built Hatch! One IDE for the full loop: explore ideas → test out designs with SuperDesign.Dev → find right skills for the job → ship with agents. > build: supports Claude code, OpenCode and Cursor agents in parallel with worktrees. > idea maze: use the canvas to sketch out ideas, paste tweets or images. > idea maze: let your agent make connections, critique your ideas or come up with related ones. > design: bring SuperDesign.Dev prompts or exports directly into your workspaces. vibedesign before vibecoding. > skills: ask your agent what skills to get with your workspace context. All open source!

Firat Sertgoz

38,578 views • 9 months ago

$FET just quietly built the internet for AI agents. 3,000,000 agents. One network. One token. Here's what nobody is explaining: Every AI agent needs 3 things to be useful: An address. A way to be found. A way to get paid. Agentverse gives agents all three. Your agent becomes discoverable on ASI: One. Other agents can find it, call it, or hire it. Automatically. No human in the loop. This is agent-to-agent commerce. And every transaction settles in $FET. More agents = more activity = more demand. The infrastructure is already live. The economy is just getting started.

$FET just quietly built the internet for AI agents. 3,000,000 agents. One network. One token. Here's what nobody is explaining: Every AI agent needs 3 things to be useful: An address. A way to be found. A way to get paid. Agentverse gives agents all three. Your agent becomes discoverable on ASI: One. Other agents can find it, call it, or hire it. Automatically. No human in the loop. This is agent-to-agent commerce. And every transaction settles in $FET. More agents = more activity = more demand. The infrastructure is already live. The economy is just getting started.

2xnmore

12,894 views • 6 months ago

This is a fun project Lakedbed, Herdr, Pi, Effect, XState prototyping a looping autonomous agent that monitors issues and clears its own backlog. The project is dogfooding an issue tracker into a durable agent that is building the issue tracker its looping over. It's an imperfect demo and the result isn't polished, but I think there's a lot fo interesting ideas to explore and expand on. The goal here is to demo a stack of tools that have been giving me consistently reliable results. Herdr and Pi make a KILLER combo for building a bespoke custom harness around your work. Lakebed is really nice to work with. Plenty of other opinions in the workshop lol - Pi as the agent harness — a minimal loop you own instead of rent, running GPT-5.5 through a Codex subscription inside a Docker workshop container - Planning before prompting: TLDraw sketches, a grilling session with Matt Pocock's Wayfinder skill, and a VISION.md to act as guardrails for the looping agent - Event sourcing with Effect v4 and Effect Schema, ports-and-adapters so the JSONL issue store can become GitHub Issues or Linear or whatever later - Multi-agent orchestration in herdr: an operator pane and an agent-loop pane, where one Pi session literally types prompts into another (fuckin LOVE this) - Enforcement that actually bites: OxLint with Ultracite, Lefthook hooks, and an agent that tries to eslint-disable its way past a complexity rule before fixing it properly 🤡 - A live issue tracker UI on Lakebed, built by the loop it tracks Workshop repo: The repo has a docker container to fire it up as a sandbox. Chapters: 0:00 Intro: Loopcraft and the workshop setup 2:25 Starting Pi inside the Docker container 4:31 Designing the looping issue resolver 10:16 Choosing XState v5 and Effect v4, prompting the starter 15:17 Why Pi: a minimal harness you own 19:35 Reviewing the generated scaffold 24:53 Architecture summary page with Lakebed 30:43 Matt Pocock skills and the vision.md concept 36:37 Grilling session: defining the vision 46:54 Daemon architecture and spec/ticket skills 55:38 Agent loop pane: one agent prompting another 1:01:22 Watching the build: compaction, tests, lint enforcement 1:07:59 Building the Loopcraft Pi extension and monitor 1:15:32 Lakebed issue tracker UI 1:21:07 Kanban board live: running the loop end to end 1:31:49 Wrap-up: bugs, next steps, takeaways

This is a fun project Lakedbed, Herdr, Pi, Effect, XState prototyping a looping autonomous agent that monitors issues and clears its own backlog. The project is dogfooding an issue tracker into a durable agent that is building the issue tracker its looping over. It's an imperfect demo and the result isn't polished, but I think there's a lot fo interesting ideas to explore and expand on. The goal here is to demo a stack of tools that have been giving me consistently reliable results. Herdr and Pi make a KILLER combo for building a bespoke custom harness around your work. Lakebed is really nice to work with. Plenty of other opinions in the workshop lol - Pi as the agent harness — a minimal loop you own instead of rent, running GPT-5.5 through a Codex subscription inside a Docker workshop container - Planning before prompting: TLDraw sketches, a grilling session with Matt Pocock's Wayfinder skill, and a VISION.md to act as guardrails for the looping agent - Event sourcing with Effect v4 and Effect Schema, ports-and-adapters so the JSONL issue store can become GitHub Issues or Linear or whatever later - Multi-agent orchestration in herdr: an operator pane and an agent-loop pane, where one Pi session literally types prompts into another (fuckin LOVE this) - Enforcement that actually bites: OxLint with Ultracite, Lefthook hooks, and an agent that tries to eslint-disable its way past a complexity rule before fixing it properly 🤡 - A live issue tracker UI on Lakebed, built by the loop it tracks Workshop repo: The repo has a docker container to fire it up as a sandbox. Chapters: 0:00 Intro: Loopcraft and the workshop setup 2:25 Starting Pi inside the Docker container 4:31 Designing the looping issue resolver 10:16 Choosing XState v5 and Effect v4, prompting the starter 15:17 Why Pi: a minimal harness you own 19:35 Reviewing the generated scaffold 24:53 Architecture summary page with Lakebed 30:43 Matt Pocock skills and the vision.md concept 36:37 Grilling session: defining the vision 46:54 Daemon architecture and spec/ticket skills 55:38 Agent loop pane: one agent prompting another 1:01:22 Watching the build: compaction, tests, lint enforcement 1:07:59 Building the Loopcraft Pi extension and monitor 1:15:32 Lakebed issue tracker UI 1:21:07 Kanban board live: running the loop end to end 1:31:49 Wrap-up: bugs, next steps, takeaways

joel ⛈️

15,967 views • 3 months ago

Semantic search alone doesn't cut it. Neither does brute-force grep. Agents need both. Today we're shipping the Retrieval Harness in LlamaParse Index: semantic search, server-side grep, and file-level navigation working together in a single agent reasoning loop. 🦙🌤️ Grep a file, list what's in an index, read past a chunk boundary, run hybrid search with reranking — all as native agent tools. Now in beta across all paid tiers. Full breakdown in the blog 👇 Learn More:

Semantic search alone doesn't cut it. Neither does brute-force grep. Agents need both. Today we're shipping the Retrieval Harness in LlamaParse Index: semantic search, server-side grep, and file-level navigation working together in a single agent reasoning loop. 🦙🌤️ Grep a file, list what's in an index, read past a chunk boundary, run hybrid search with reranking — all as native agent tools. Now in beta across all paid tiers. Full breakdown in the blog 👇 Learn More:

LlamaIndex 🦙

29,181 views • 3 months ago

Introducing Roughdraft! A new open source project designed to make collaboration with agents better. The idea is to bring commenting and suggested changes to markdown (e.g. plan docs) in a nice interface. Free, local, etc. 👉 👈

Introducing Roughdraft! A new open source project designed to make collaboration with agents better. The idea is to bring commenting and suggested changes to markdown (e.g. plan docs) in a nice interface. Free, local, etc. 👉 👈

Nathan Baschez

401,696 views • 5 months ago

Loop engineering is great until something breaks. Here is how I improve the reliability of my agentic loops. I use human-in-the-loop (HITL). It's easy and extremely effective. Anyone can build this. My setup: I recorded a quick demo of how it all works. I shared recently that I now use more voice agents to build and communicate with agents. I also use them to verify. I hate the idea of being tied down to my computer or in a Slack channel to communicate with my agents. Here is what I have done to streamline communication with my agents. All my Claude and Codex agent sessions now use the Dial MCP server. It has a bunch of tools and provisions my agents with their own number that can place calls as native tools, with voice, SMS, and iMessage behind one interface. As my loops/automations work on PRs and new features, my agents escalate decisions to me via a short phone call. This is extremely useful when I am on the road or away from my desk. If you want to try this with Claude Code or Codex, paste this into your agent and get started right away: "Get yourself a Dial phone number and call me. Say hello and that setup is working, then hang up. Follow Natan Voitenkov and team are building something special here. Go check them out. Give your agent a phone number now: ($5 free credit)

Loop engineering is great until something breaks. Here is how I improve the reliability of my agentic loops. I use human-in-the-loop (HITL). It's easy and extremely effective. Anyone can build this. My setup: I recorded a quick demo of how it all works. I shared recently that I now use more voice agents to build and communicate with agents. I also use them to verify. I hate the idea of being tied down to my computer or in a Slack channel to communicate with my agents. Here is what I have done to streamline communication with my agents. All my Claude and Codex agent sessions now use the Dial MCP server. It has a bunch of tools and provisions my agents with their own number that can place calls as native tools, with voice, SMS, and iMessage behind one interface. As my loops/automations work on PRs and new features, my agents escalate decisions to me via a short phone call. This is extremely useful when I am on the road or away from my desk. If you want to try this with Claude Code or Codex, paste this into your agent and get started right away: "Get yourself a Dial phone number and call me. Say hello and that setup is working, then hang up. Follow Natan Voitenkov and team are building something special here. Go check them out. Give your agent a phone number now: ($5 free credit)

elvis

15,614 views • 3 months ago

Andrej Karpathy just broke the entire premise of modern AI: "The new computer isn't a chip - It's a language model running agent loops." Agents talking to agents -loops spawning sub-loops The stack being built RIGHT NOW: agent loop + memory + sub-agents + harness In a 20-minute talk, Karpathy explain the full reasoning stack

Andrej Karpathy just broke the entire premise of modern AI: "The new computer isn't a chip - It's a language model running agent loops." Agents talking to agents -loops spawning sub-loops The stack being built RIGHT NOW: agent loop + memory + sub-agents + harness In a 20-minute talk, Karpathy explain the full reasoning stack

Ricker

32,015 views • 3 months ago

A project brief should live with the team, not in one person's AI chat. That's the Humans + Agents idea behind Teamily AI: the people and the agents on a project work from one shared brief in one group, and the deliverable stays open for all of them to keep building on. Teamily AI To test that, I published a Loop in my Ad Creative Studio group: the Content Creator agent turns customer complaints into an onboarding campaign, then hands it to me for review. The output is an Artifact: "From complaint to campaign: the onboarding loop, before & after". The screen proof is the handoff, end to end: research, strategy, content, a reviewer challenging weak claims, then the same loop rerun with the audience flipped from individual users to small teams. Then a partner I'm working with joined the group and asked to rerun the same loop for small teams instead of individual users. The update landed in the same Artifact, and the audience, requirements and earlier feedback stayed with the project as it developed. Teamily AI is a human-agent collaborative workspace, great for small businesses and teams. The finished Artifact: Code: AELNQ9

A project brief should live with the team, not in one person's AI chat. That's the Humans + Agents idea behind Teamily AI: the people and the agents on a project work from one shared brief in one group, and the deliverable stays open for all of them to keep building on. Teamily AI To test that, I published a Loop in my Ad Creative Studio group: the Content Creator agent turns customer complaints into an onboarding campaign, then hands it to me for review. The output is an Artifact: "From complaint to campaign: the onboarding loop, before & after". The screen proof is the handoff, end to end: research, strategy, content, a reviewer challenging weak claims, then the same loop rerun with the audience flipped from individual users to small teams. Then a partner I'm working with joined the group and asked to rerun the same loop for small teams instead of individual users. The update landed in the same Artifact, and the audience, requirements and earlier feedback stayed with the project as it developed. Teamily AI is a human-agent collaborative workspace, great for small businesses and teams. The finished Artifact: Code: AELNQ9

Sean Wallace

53,300 views • 11 days ago