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.

pi-web-access update: the agent can now search, then pull specific values off a page instead of reading the whole thing. Clairic returns just what you asked for, with the line it came from, and flags anything it isn't sure of. When in codemode the values come back as typed... show more

Nico Bailon

8,972 subscribers

18,145 views • 6 days ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

14 Comments

Kipoo6 days ago

nice job，我觉得pi-web-access可以成为pi的内置插件

.6 days ago

web access and generally speaking, file/computer access, should be a default functionality in Pi agent

Ruben6 days ago

Thank you @nicopreme. What resources would you reccomend for understanding how agents use web search and its different variations? I could not find any comprehensive review.

Diego Cerdan6 days ago

Thank you for the Pi extension. Looking good! I will try it out and maybe replace my current custom SearXNG skill.

HM6 days ago

Keeping the source line is the good part. For prices, I'd want the billing period and 'starting at' qualifier attached too. A clean number can be misleading on its own.

Sandip Bhattacharya 🌏✌️6 days ago

Is this an exact string/regex match on the content? or a subagent based semantic extraction?

kayce6 days ago

有点像codemode了

Veyro6 days ago

The uncertainty flag is the part I'd want to keep in the typed result too. A script needs to distinguish “not found” from a confidently extracted zero.

Niko Petrov6 days ago

Cutting tool tokens by 60% is huge, context window was killing us.

Jonathan Vital6 days ago

What a gain!!

Ravi6 days ago

an agent that grabs the one value it needs and shows the line it came from already reads docs better than most of us

Web制作出身のAI・Webアプリ開発者5 days ago

全文を渡さず、必要な値と根拠の行だけを返し、不確かな箇所も示す設計は、調査の速さと検証しやすさを両立できますね。エージェントに任せる範囲を広げても出典を追えるのは、実務で大事だと思います。

Peter Dove6 days ago

Nice. Targeted extraction should cut token waste and simplify price scripts

BreezeOg6 days ago

does it ever pull the wrong value and still show a source

Related Videos

pi-web-access is an extension that provides web search (and more) for Pi coding agent. The latest update introduces a new UX pattern: you can hit `Ctrl+S` to intercept agent searches mid-flight to inspect and curate exactly what makes it into the agent conversation. pi install npm:pi-web-access

pi-web-access is an extension that provides web search (and more) for Pi coding agent. The latest update introduces a new UX pattern: you can hit `Ctrl+S` to intercept agent searches mid-flight to inspect and curate exactly what makes it into the agent conversation. pi install npm:pi-web-access

Nico Bailon

29,352 views • 8 months ago

Updated pi-web-access for Pi agent with video understanding. If you're signed into Gemini in Chrome, no API keys needed for web search or video analysis. Point it at any YouTube URL or local video file (mp4, mov, etc) and ask questions about the content. pi install npm:pi-web-access

Updated pi-web-access for Pi agent with video understanding. If you're signed into Gemini in Chrome, no API keys needed for web search or video analysis. Point it at any YouTube URL or local video file (mp4, mov, etc) and ask questions about the content. pi install npm:pi-web-access

Nico Bailon

18,367 views • 9 months ago

Introducing AgentKey by Chainbase Labs. AgentKey connects AI agents to the outside world. With one install, your agent can access web search, read pages, reach social platforms, query crypto data, and more. Try it now:

Introducing AgentKey by Chainbase Labs. AgentKey connects AI agents to the outside world. With one install, your agent can access web search, read pages, reach social platforms, query crypto data, and more. Try it now:

Chainbase (💜,💛)

23,102 views • 5 months ago

addicted to using boomerang mode (aka reverse D-Mail) in Pi these days. ctrl+alt+b enables it for the next prompt submitted -> after the prompt runs, it rewinds back to the same point with file changes intact + leaves a summary in the feed so the agent knows what happened. using it often can make the context window feel nearly unlimited. Powered by the native /tree functionality in pi. pi install pi-boomerang

addicted to using boomerang mode (aka reverse D-Mail) in Pi these days. ctrl+alt+b enables it for the next prompt submitted -> after the prompt runs, it rewinds back to the same point with file changes intact + leaves a summary in the feed so the agent knows what happened. using it often can make the context window feel nearly unlimited. Powered by the native /tree functionality in pi. pi install pi-boomerang

Nico Bailon

38,562 views • 6 months ago

We gave Pi Code persistent memory with Mem0 Launching the Mem0 plugin for Pi Code: persistent, scoped, semantic memory across sessions and projects. It captures what matters, searches by meaning, and brings the right context back when your agent needs it. Try it out: pi install npm:@ mem0/pi-agent-plugin

We gave Pi Code persistent memory with Mem0 Launching the Mem0 plugin for Pi Code: persistent, scoped, semantic memory across sessions and projects. It captures what matters, searches by meaning, and brings the right context back when your agent needs it. Try it out: pi install npm:@ mem0/pi-agent-plugin

mem0

25,165 views • 4 months ago

🚨 THIS IS ACTUALLY INSANE Your AI agent can have access to the web. But if it can't reliably read what’s actually on the page, that access is almost useless. We looked at Firecrawl as the web layer for AI agents and the numbers are hard to ignore. The setup is simple: Give it a URL, search query, or website. Firecrawl handles the ugly part — crawling, scraping, rendering, extracting, and turning web content into something an AI model can actually use. The headline numbers: → 173,000+ GitHub stars → Search, scrape and interact with the web at scale → Supports web pages, PDFs, DOCX and other content → Structured data extraction for AI workflows → MCP support for connecting it directly to AI agents The workflow looks like this: Search → Scrape → Crawl → Extract → Feed the agent Three things stand out: 1. Scraping becomes an infrastructure layer Instead of maintaining your own pile of HTTP clients, parsers, browser automation and retry logic, you can treat web access as an API. 2. Agents get more than raw HTML The goal isn't just downloading a webpage. It's turning messy web content into clean context that an LLM can reason over. 3. The same layer works across different agent workflows Research agents. RAG pipelines. AI search. Competitive intelligence. Web-data extraction. The interesting shift: AI agents don't just need better models. They need better access to the information those models are supposed to reason about. Firecrawl is building that layer. Save this repo.

🚨 THIS IS ACTUALLY INSANE Your AI agent can have access to the web. But if it can't reliably read what’s actually on the page, that access is almost useless. We looked at Firecrawl as the web layer for AI agents and the numbers are hard to ignore. The setup is simple: Give it a URL, search query, or website. Firecrawl handles the ugly part — crawling, scraping, rendering, extracting, and turning web content into something an AI model can actually use. The headline numbers: → 173,000+ GitHub stars → Search, scrape and interact with the web at scale → Supports web pages, PDFs, DOCX and other content → Structured data extraction for AI workflows → MCP support for connecting it directly to AI agents The workflow looks like this: Search → Scrape → Crawl → Extract → Feed the agent Three things stand out: 1. Scraping becomes an infrastructure layer Instead of maintaining your own pile of HTTP clients, parsers, browser automation and retry logic, you can treat web access as an API. 2. Agents get more than raw HTML The goal isn't just downloading a webpage. It's turning messy web content into clean context that an LLM can reason over. 3. The same layer works across different agent workflows Research agents. RAG pipelines. AI search. Competitive intelligence. Web-data extraction. The interesting shift: AI agents don't just need better models. They need better access to the information those models are supposed to reason about. Firecrawl is building that layer. Save this repo.

Vikas gupta

20,121 views • 1 month ago

Oxylabs Web API x Muse. Connect Web API to Meta’s personal AI agent and let it search and pull fresh data from any public page, on a schedule you set. Track which AI agent tools the top search results recommend, with a new report every morning, even after you close the app!

Oxylabs Web API x Muse. Connect Web API to Meta’s personal AI agent and let it search and pull fresh data from any public page, on a schedule you set. Track which AI agent tools the top search results recommend, with a new report every morning, even after you close the app!

Oxylabs

31,697 views • 11 days ago

A screen recording is the best context you can give an AI agent. It's also the worst. A few seconds of video is thousands of images. Feed that to an agent and you instantly blow up its memory. So I built a tool that fixes it. How it works: Hit record and just do your thing. Click around, type, and talk out loud about what you're doing. While you record, it quietly captures the stuff that actually matters: your voice transcription, every click and the exact thing you clicked, what you typed, and the shortcuts you press. When you stop, it doesn't hand the agent the whole video. It writes a short, readable summary: a timeline of what you said and did. The agent reads that tiny summary first. Then, only when it needs to actually see a moment, it pulls that one exact frame from the video, like flipping to a single page instead of reading the whole book. The result: the agent understands everything you did! Full context. Less tokens.

A screen recording is the best context you can give an AI agent. It's also the worst. A few seconds of video is thousands of images. Feed that to an agent and you instantly blow up its memory. So I built a tool that fixes it. How it works: Hit record and just do your thing. Click around, type, and talk out loud about what you're doing. While you record, it quietly captures the stuff that actually matters: your voice transcription, every click and the exact thing you clicked, what you typed, and the shortcuts you press. When you stop, it doesn't hand the agent the whole video. It writes a short, readable summary: a timeline of what you said and did. The agent reads that tiny summary first. Then, only when it needs to actually see a moment, it pulls that one exact frame from the video, like flipping to a single page instead of reading the whole book. The result: the agent understands everything you did! Full context. Less tokens.

Milind S

12,486 views • 4 months ago

Agents perform better when they can take action, not just think! kevin from Coinbase explains why x402 matters. As it enables an open web of tool calls and realtime data access, so agents can actually execute on tasks instead of operating in isolation.

Agents perform better when they can take action, not just think! kevin from Coinbase explains why x402 matters. As it enables an open web of tool calls and realtime data access, so agents can actually execute on tasks instead of operating in isolation.

SKALE

54,102 views • 8 months ago

Here is why Nimble has 1.1M Claude plugin installations. Ask Claude what your competitor's pricing looks like right now and it will tell you it can't browse the web. Install the Nimble skill and ask the same question. You get live, structured data back in seconds. That's the whole story. Nimble turns Claude into an agent that can actually search the live internet, extract specific data from any page, and return it in clean, structured format ready for analysis. No setup beyond the one-time install. → Search any public webpage from inside Claude → Extract structured data from product pages, news, job boards, social profiles → Get AI-ready output, not raw HTML It crossed 1.1M installs because it solves the exact limitation that makes Claude frustrating for real research tasks. Link:

Here is why Nimble has 1.1M Claude plugin installations. Ask Claude what your competitor's pricing looks like right now and it will tell you it can't browse the web. Install the Nimble skill and ask the same question. You get live, structured data back in seconds. That's the whole story. Nimble turns Claude into an agent that can actually search the live internet, extract specific data from any page, and return it in clean, structured format ready for analysis. No setup beyond the one-time install. → Search any public webpage from inside Claude → Extract structured data from product pages, news, job boards, social profiles → Get AI-ready output, not raw HTML It crossed 1.1M installs because it solves the exact limitation that makes Claude frustrating for real research tasks. Link:

Rimsha Bhardwaj

61,011 views • 5 months ago

Big one for the Hermes Agent community. TinyFish is now a plugin in the Nous Research Hermes Agent. Your Hermes agent gets free Search and Fetch, so it can find and read the live web at $0. When a task needs real clicks, TinyFish Browser and Web Agent are one approval away.

Big one for the Hermes Agent community. TinyFish is now a plugin in the Nous Research Hermes Agent. Your Hermes agent gets free Search and Fetch, so it can find and read the live web at $0. When a task needs real clicks, TinyFish Browser and Web Agent are one approval away.

TinyFish

48,494 views • 23 days ago

It takes 5 minutes to create a world-class web research agent that runs 24/7 and returns structured search results. 1. You can limit the sources it's allowed to use 2. You can specify the goals it needs to complete 3. You can design the format structure of the results 4. You can configure how much computing power it will use I show you how to set up a Web Search Agent in the video. A few benefits here: 1. This is great for complex research tasks 2. It's really easy to add this like a LEGO piece to a more complex workflow 3. Create an agent once, and reuse it with different inputs 4. The agent adapts to your use case as you work with it By the way, Nimble published a whole library of templates and research agents you can reuse and modify. You can start sending queries to one of these existing agents, and it will begin to adapt to Go here to try it out: Thanks to the Nimble team for partnering with me on this post.

It takes 5 minutes to create a world-class web research agent that runs 24/7 and returns structured search results. 1. You can limit the sources it's allowed to use 2. You can specify the goals it needs to complete 3. You can design the format structure of the results 4. You can configure how much computing power it will use I show you how to set up a Web Search Agent in the video. A few benefits here: 1. This is great for complex research tasks 2. It's really easy to add this like a LEGO piece to a more complex workflow 3. Create an agent once, and reuse it with different inputs 4. The agent adapts to your use case as you work with it By the way, Nimble published a whole library of templates and research agents you can reuse and modify. You can start sending queries to one of these existing agents, and it will begin to adapt to Go here to try it out: Thanks to the Nimble team for partnering with me on this post.

Santiago

21,079 views • 3 months ago

Your app can now search the web for images. Web search in the Responses API now supports image results in addition to text results, so you can build apps that surface products, places, visual references, and source links for inspiration.

Your app can now search the web for images. Web search in the Responses API now supports image results in addition to text results, so you can build apps that surface products, places, visual references, and source links for inspiration.

OpenAI Developers

155,571 views • 4 months ago

Cursor now has a built-in web browser. You can say "start my app" and it will run the dev server in a terminal and then navigate the browser to localhost:3000. It can take screenshots of the page for iterating on design, and you can send DOM elements to the agent with prompts!

Cursor now has a built-in web browser. You can say "start my app" and it will run the dev server in a terminal and then navigate the browser to localhost:3000. It can take screenshots of the page for iterating on design, and you can send DOM elements to the agent with prompts!

Lee Robinson

216,307 views • 1 year ago

Stop paying $20–50/month for your AI agent to browse the web. i built the free version no API keys, no config, no rate limits😳 it's called Hound. An MCP server that does fetch, search, crawl, and screenshot — all keyless and now it ships a native Pi extension, so you get all 6 tools as first-class Pi tools, not through a generic MCP adapter what you get: → web_fetch — anti-bot fetch → clean markdown. PDFs get section maps + auto-OCR. Dead pages auto-recover from the Internet archive (honestly marked, never pretending it's live) → web_search — 10 keyless backends in parallel (DuckDuckGo, Brave, Mojeek, Yandex, Google…) + a local neural reranker → web_crawl — best-first same-domain walk. sitemap mode maps a whole site in one fetch → web_screenshot — anti-bot shots for multimodal models → cache_clear + hound_version what it replaces: -tavily — free tier -firecrawl — $16–50/mo -searXNG + browser + extractor glue — fragile -hound — $0, keyless, forever why it matters: >agent research needs web access, but everyone gates it behind a key or a limit >hound is honest about errors — a 404 says "Page doesn't exist" instead of dumping an error page as if it were content >archive recovery tags source= so the agent knows the content is historical, not live 2-minute setup: -pip install hound-mcp[all] -pi install npm: @houndmcp/hound-mcp-pi -/reload → tools are live. That's it Caveats (read): dataDome, Akamai, Cloudflare Turnstile — no free tool bypasses these hound tries the stealthy browser, then tells you to switch sources the [all] extra is ~100MB (onnxruntime + OCR). Works without it, but full install is recommended not a scraping-at-scale tool. Crawl caps at 100 pages by default — built for agent research Give your agent the web for free

Stop paying $20–50/month for your AI agent to browse the web. i built the free version no API keys, no config, no rate limits😳 it's called Hound. An MCP server that does fetch, search, crawl, and screenshot — all keyless and now it ships a native Pi extension, so you get all 6 tools as first-class Pi tools, not through a generic MCP adapter what you get: → web_fetch — anti-bot fetch → clean markdown. PDFs get section maps + auto-OCR. Dead pages auto-recover from the Internet archive (honestly marked, never pretending it's live) → web_search — 10 keyless backends in parallel (DuckDuckGo, Brave, Mojeek, Yandex, Google…) + a local neural reranker → web_crawl — best-first same-domain walk. sitemap mode maps a whole site in one fetch → web_screenshot — anti-bot shots for multimodal models → cache_clear + hound_version what it replaces: -tavily — free tier -firecrawl — $16–50/mo -searXNG + browser + extractor glue — fragile -hound — $0, keyless, forever why it matters: >agent research needs web access, but everyone gates it behind a key or a limit >hound is honest about errors — a 404 says "Page doesn't exist" instead of dumping an error page as if it were content >archive recovery tags source= so the agent knows the content is historical, not live 2-minute setup: -pip install hound-mcp[all] -pi install npm: @houndmcp/hound-mcp-pi -/reload → tools are live. That's it Caveats (read): dataDome, Akamai, Cloudflare Turnstile — no free tool bypasses these hound tries the stealthy browser, then tells you to switch sources the [all] extra is ~100MB (onnxruntime + OCR). Works without it, but full install is recommended not a scraping-at-scale tool. Crawl caps at 100 pages by default — built for agent research Give your agent the web for free

Nahid

45,466 views • 3 months ago

New Pi extension: pi-cached-compaction. Pi's built-in compaction swaps the system prompt, drops the tools, and flattens the messages, which breaks the prompt cache. This one keeps all three as they are and just appends the summary request, so ~99% is read from cache. And if you step away, it compacts just before the cache expires, so your next message doesn't pay to cache the whole session again. pi install npm:pi-cached-compaction

New Pi extension: pi-cached-compaction. Pi's built-in compaction swaps the system prompt, drops the tools, and flattens the messages, which breaks the prompt cache. This one keeps all three as they are and just appends the summary request, so ~99% is read from cache. And if you step away, it compacts just before the cache expires, so your next message doesn't pay to cache the whole session again. pi install npm:pi-cached-compaction

Nico Bailon

20,663 views • 10 days ago

Without ever leaving Databricks, you can search the web and pull live, structured data into your tables in plain english by using Genie to call Nimble. Step 1: Search the web and structure the data ⬇️

Without ever leaving Databricks, you can search the web and pull live, structured data into your tables in plain english by using Genie to call Nimble. Step 1: Search the web and structure the data ⬇️

Nimble

51,119 views • 4 months ago

$PI is upgrading to Protocol 28. If it sounds familiar, that’s because it is. Pi is built on Stellar. Pi Network’s just upgraded its testnet to Protocol 28. If you follow my stuff, these features might sound familiar. There’s a good reason for that, and it’s worth knowing. So what does it do? It improves how the network handles delays in transaction data, it lets developers upgrade whole groups of smart contracts at once, and it lets them change stored data more safely as their apps evolve. If you saw my Stellar video, that’s the exact same upgrade Stellar just did. That’s not a coincidence. $PI is actually built on Stellar’s technology. So when $XLM ships a protocol upgrade, Pi inherits the same one. That’s why they line up. As for the timing, node operators need to upgrade by October 13th, and Pi’s planning to bring it to the mainnet on October 16th. Nothing dramatic. Just Pi keeping pace with the tech it’s built on.

$PI is upgrading to Protocol 28. If it sounds familiar, that’s because it is. Pi is built on Stellar. Pi Network’s just upgraded its testnet to Protocol 28. If you follow my stuff, these features might sound familiar. There’s a good reason for that, and it’s worth knowing. So what does it do? It improves how the network handles delays in transaction data, it lets developers upgrade whole groups of smart contracts at once, and it lets them change stored data more safely as their apps evolve. If you saw my Stellar video, that’s the exact same upgrade Stellar just did. That’s not a coincidence. $PI is actually built on Stellar’s technology. So when $XLM ships a protocol upgrade, Pi inherits the same one. That’s why they line up. As for the timing, node operators need to upgrade by October 13th, and Pi’s planning to bring it to the mainnet on October 16th. Nothing dramatic. Just Pi keeping pace with the tech it’s built on.

Rypto

19,653 views • 1 month ago