Timeline

Speaker 1: Gotta give that a test run. Speaker 2: Yeah, I'm gonna see if I can do it right now. I'll see. Speaker 1: All right. While you're looking, anyone else got some, any good news, any updates? Speaker 3: Me and Merlino, me and Merlino got an agency together. Speaker 1: Oh yeah?

Speaker 1: I'll crawl that board then. Speaker 2: Gotta get it right, I'm sure. Speaker 1: I'm gonna see if I can get it right now. I'll see. Speaker 2: Alright, while you're looking, anyone else got any good news, any updates? Speaker 3: I have a question, what will... Speaker 4: I haven't seen Marlena. Me and Marlena got agency.

[Web Browser Content] HercBrain — Mike Merlino Agent Operating System Features Vault SUPERMEMORY Ask Gemini Work Activate hercbrain PRIVATE AGENT OPERATING SYSTEM POWERED BY HERMES + SUPERMEMORY The agent brain that knows what matters Workflow Security FAQ HERC RUNTIME VAULT

Good work. I haven't tried it in a second. But yeah, it's just crawled out of orbit.

[Chat Transcript] Unigram Matteo: Your version is built as a standalone static site... Ava: my flights for Austin texas MacHermes: I made your version of agentbrain.sh as HercBrain and deployed it. Live Vercel URL: https://herc-agentbrain-mike.vercel.app Verified: - Site loads: 200 OK - CSS loads: 200 OK - JS loads: 200 OK - Private/noindex header active: X-Robots-Tag: noindex, nofollow, noarchive - Content markers verified: HercBrain, Private Agent Operating System, second-brain / Supermemory / Hermes positioning. What I adapted from the reference: - Warm editorial SaaS style - Large serif headline with italic accent words - Cream paper background - Thin rule dividers - Numbered capability rows - Enterprise/security-style inset section - FAQ accordion - Dark footer - Minimal premium CTAs - Light default with always-visible dark toggle What I made it about: - Mike/Herc operating brain - Hermes + Supermemory - /data/mike-second-brain - SEO/AI visibility reports - exact-match domain leadgen - one-link Vercel deliverables - private command-center workflow HercBrain — Mike Merlino Agent Operating System A private operating system for Mike Merlino: Hermes agent runtime, Supermemory, second-brain vault, SEO visibility reports, domains leadgen, and Vercel delivery.

Screen: Creatif o Code Adminj unclecode Issues A Transla A Transla github.com/unclecode/craw14ai craw14ai Transla Transla Discussions Q Transla Actions Go to file A Merline Projects Merline Plugins Merline + 2 Q Type LI to search Ask Gemini Star 65.6k o Work x 28 Pull requests Public 7 Branches 61 Agents Tags Security and quality 2 Watch 362 Sponsor L Insights craw14ai main unclecode .claude .context .github craw14 Fork 6.7k About Merge pull request #1885 from unclecode/develop Add file Idebe5f • 2 months ago < > Code O 1,468 Commits 2 months ago 2 months ago 2 months ago 2 months ago 2 months ago 2 months ago last year 4 months ago 3 months ago 2 months ago Craw14Al: Open-source LLM Friendly Web Crawler & Scraper. Don't be shy, join here: https://discord.gg(jP8KfhDhyN test: add comprehensive regr

Speaker 1: So, uh, Andrew's local AI scraper? Yeah, it's something we built for the AI overview in the local. I don't know if he posted that or not. I don't know if he's here. Speaker 2: Yeah, I'm here. Hey, dude. Speaker 1: There he is. Yeah, it should work. I haven't tried it in a second, but it's just...

[Chat Transcript] Zoom Meeting Participants: Rich Fong Chris Labbate Michael Merlino Chris Castillo Bret Acott Meeting controls: - Recording and taking notes - Richard's Fathom Notetaker - AI Companion Status: This meeting is being transcribed.

Speaker 1: Uh, yeah. Yeah, did anybody try out, uh, Andrew's local AI scraper? He had something going for the AI overviews in local. I don't know if he posted that or not. I don't know if he's here. Speaker 2: Yeah, I'm here. Speaker 1: There he is. Speaker 2: Yeah, I should...

Speaker 1: Yeah, how much is that not gonna be for scraping when you could just have that shit running on a fucking something you probably already got running? There's a lot of local free models, bro, that get busy. You know what I'm saying? Uh, real busy. Just throwing that in the air. Speaker 2: Yeah, did anybody try...

...something scraped immediately and I'm doing it via Firecrawl and there's other ways but Firecrawl gets every motherfucking thing that I want possibly and my agent is trained on that, you know what I mean? Like, because if you're going to start paying, paying, if you think about it, at the end of the year...

We could try Hermes, yeah.

Speaker 1: Unless you're really going for speed, don't pay to scrape, bro. Get a fucking local model going that's way capable of doing that for free on a server or on something, you know? Speaker 2: You could try Hermes, yeah. Speaker 1: Uh, no, not try. I do not pay to scrape, bro, unless I want—

[Command-Line Output] Claude Code v2.1.142 Opus 4.7 (1M context) Working Directory: D:\ClaudeDev UserPromptSubmtt says: view your session in honcho GUI: [honcho] user-prompt injected conclusions (cached) Skill (/install) Successfully loaded skill https://github.com/unclecode/crawl4ai https://app.honcho.dev/explore?workspace=claude_code&view=sessions&session=mtke-claudedev 11 MCP servers need auth /mcp

Screen: ORESTOCKO MacHermes bot terminal: "set -euo pipefail URL='https://merlin... May 14 todo: "updating 1 task(s)" agentbrain.sh AgentBrain — Enterprise agent runtime The agent runtime that knows who can do what. Enterprise-grade RBAC, audit, and on-prem deployment, powered by GoClaw. 11:07 PM Got it — making a Mike/Merlino version of AgentBrain.sh : same vibe/structure, but original copy/branding for your ecosystem. I'll build and deploy a Vercel preview. 11:07 PM Unread Messages I've got the reference: beige editorial SaaS landing page, black text, massive type, numbered cards, compatibility grid, security/ method/FAQ/CTA. Building yours now. 11:07 PM Deploy tripped on the reference screenshot binary. Skipping that asset and redeploying the actual site. 11:09 PM Vercel made it private by defa

to scrape like that, it was about 15 bucks towards my max Claude AI subscription. Uh, also OpenClaw, I think got back into the subscription. Yeah, go ahead.

[Chat Transcript] Unigram Recent Activity: - Matteo: Your version is built as a standalone static site... - Herc: I made your version of agentbrain... - MacHermes: Done — your AgentBrain-style version is live: https://merlino-agentbrain-nxgk46pa9-mmerlin023s-projects.vercel.app - Herman (Bot): Lead Engine 20260515T140854Z Scored: 69... - Ava: my flights for Austin texas - Oscar (Bot): make my version https://agentbrain.sh/ System Status: - ClickUp snapshot: No overdue tasks. - Fireflies: Could not verify May 14 action items. - Cron: Creation blocked due to Gateway TransportError. Priorities: 1. Reply to Oxylabs. 2. Decide on Google Cloud billing escalation. 3. Protect 11:30-3:00 calendar block. Lead Engine Report (May 15, 2026): - Campaign batch: 48 - Top leads include MasterTech Air, Tops Air Conditioning, and Blue Valley Electric. SEO Keywords: - Best SEO in Denver, How Much Does a Website Cost in Sarasota, Google Ads Management Sarasota, Video Marketing Sarasota, Is Local SEO Worth It in 2026, Ranking Reels, SEO in NYC.

Speaker 1: When you do a scrape like that, it was about 15 bucks towards my max Claude AI subscription. Uh, also OpenClaw [?], I think, got, uh, back... Speaker 2: Sorry Chris, real quick. Unless you're going for speed, my brother, like...

[Command-Line Output] North Valley Claude Code v2.1.142 Opus 4.7 (1M context) Path: D:\ClaudeDev URL: https://github.com/unclecode/craw14a Commands: - permissions (shift+tab) - install (shift+tab) - Press up to edit queued messages Skedaddling...

Speaker 1: Is it able to scrape emails and like the actual... Speaker 2: Exactly. Find the LinkedIn, all their email, telephone, website, you know, contact data. Pretty flawless from start to finish. So, wasn't too bad. It was about...

Speaker 1: ...actual— Speaker 2: Yeah, exactly. Find the info. So LinkedIn, all their email, telephone, website, you know, contact data. Uh, pretty flawless from start to finish. So, uh, wasn't too bad. It was about... when you...

[Command-Line Output] North Valley The pattern that's working: - Any visitor typing the original site URL globally bounced to /locations/ to pick a city. - Any visitor inside a silo using the original NVSP URL convention 308'd to the canonical city silo URL. - All 9 city pages x 15 cities = 135 actual content pages, all live. So every URL Mike's prospects/old GBP links/Google index might point to is now a clean 200 (after at most one redirect hop) on the new site. [honcho] response: saved response (1665 chars) Stop says: Baked for 56s. No, those would need to be on every silo individually. Imagine taking one location website, cloning it onto a subdomain and then doing that per location, but I don't want 15 different domains. I want the silo to be a website completely on its own. Create 15 pages for each of the URLs I am dropping. [honcho] user-prompt injected conclusions UserPromptSubmit says: You're right. 3 of those URLs are currently redirects instead of dedicated pages, and they need to be standalone pages on each of the 15 silos. That's 45 missing pages. The gap: - 3 URLs currently redirecting need their own pages x 15 = 45 pages. - Plus, the original NVSP canonical URL structure should be the source of truth. So I'm flipping the redirects: shortcut URLs will redirect to the original-NVSP-style canonical URLs. Agent (Build 45 missing silo pages + flip canonicals) Done (60 tool uses, 197.7k tokens, 17m 1s) 595 routes live on https://nvsp-rebuild.vercel.app. All 45 new pages built (3 types x 15 cities). The 9 URLs you listed are now all dedicated pages per city. URL structure now mirrors the original NVSP inside every silo.

[Web Browser Content] Crawl4AI: Open-source LLM Friendly Web Crawler & Scraper Repository: github.com/unclecode/crawl4ai Owner: unclecode Features: - Open-source web crawler and scraper - LLM friendly - Join the community: https://discord.gg/jP8KfhDhyN Repository Stats: - Stars: 65.6k - Forks: 6.7k - Watching: 362 - Commits: 1,468 Recent Activity: - Merge pull request #1885 from unclecode/develop - Added comprehensive regression test suite (291 tests) - HTTP strategy detects and saves file downloads - Batch merge of community PRs - Updated version references to 0.8.6 - Added network request and console message capture - Added CycloneDX SBOM and generation script

[Web Browser Content] Google Search: crawlfor ai GitHub: Craw4AI - Open-source LLM Friendly Web Crawler Reliable, large-scale web extraction, built to be cost-effective. Apply for early access. Craw4AI Documentation: https://docs.craw4ai.com Home - Craw4AI Documentation (v0.8.x) Delivers fast, AI-ready web crawling tailored for large language models, AI agents, and data pipelines. Fully open source and flexible. YouTube: The Fastest Python Scraper for RAG? (Craw4AI) Craw4AI is built for AI, outputs clean markdown/JSON, runs async, and is up to 6x faster than traditional scrapers.

[Web Browser Content] Google Search Shortcuts: - Claude (MCP) - Developers - Call Tracking - Gmail - Call Log - Legiit - My Drive - Google Drive - Website Masterlist 2025 - CTR Geeks - Ads Manager - Merlino Marketing - Facebook profile optimization Features: - Ask Google - Ask Gemini - Customize Chrome

Speaker 1: ...equivalent of like, you know, all the old guys that you used to use for basically getting leads scraping and stuff like that. I forget the guys we used to use, but it's pretty solid. Speaker 2: Okay, cool. Is it able to scrape emails and like the...

[Web Browser Content] Translation Bridge: Switch WordPress Page Builders in Seconds One API translates content across Elementor, Divi, Bricks, Gutenberg, and 5 more builders. Keep your design, change your engine. Public Beta: 9 builders supported Switch WordPress page builders in seconds, not weeks. Run Your First Translation Free POST /api/translate-and-publish See How It Works Features | Pricing | FAQ | Ask Gemini | Get API Key

Speaker 1: So we could just create leads like crazy. Speaker 2: Oh yeah? Speaker 1: It's the equivalent of like, you know, all the old guys that we used to use for basically getting leads, scraping and stuff like that. I think it's better than the guys we used to use, but it's pretty solid. Speaker 2: Okay, cool.

[Web Browser Content] EIN Presswire Pricing Package Options: 1. Get Recognized: $149 for 1 news release (365 days, $149/release). 2. Expand Your Reach: $499 for 5 releases + 1 free (365 days, $83.17/release). 3. Worldwide Attention: $999 for 15 releases (365 days, $66.60/release). Distribution includes: AP News, Nexstar Media Group, USA TODAY Network, Bloomberg Terminals, MuckRack, Google News, Google Search, Bing, and AI/LLM platforms. Terms & Conditions last updated: September 30, 2025.

[Web Browser Content] Plugins - WordPress MainWP: Manage sites from one central location. Plugin documentation and options can be found here https://mainwp.com/kb. Version 6.0.11 | By MainWP SSL Insecure Content Fixer: Clean up WordPress website HTTPS insecure content. Version 2.7.2 | By WebAware Turbo-Subdomains: Powerful sub-domain networks (PBN), that actually work. Version 4.2.148.2 | By Turboware. (Update available to 4.2.148.4) WP Reset PRO: Easily undo any change on the site by restoring a snapshot, or reset the entire site. Version 6.13 | By WebFactory Ltd WP-Sweep: Allows you to clean up unused, orphaned and duplicated data in your WordPress. Version 1.1.8 | By Lester 'GaMerZ Chan

[Web Browser Content] General Settings - WordPress Howdy, NewsJive Site Title: NewsJive WordPress Address (URL): https://newsjive.net Site Address (URL): https://newsjive.net Administration Email Address: mepeymike@tsd.guru (Pending change to: mike@merlinomarketing.com) Membership: Anyone can register New User Default Role: Subscriber Site Language: English (United States) Timezone: UTC+0 Date Format: May 15, 2026 Time Format: 3:35 pm Settings saved.

Speaker 1: Today, uh, we used Crawl4AI. Uh, there's a GitHub on that and we put together a solid, uh, web-based scraper, so we can scrape leads like crazy. Speaker 2: Oh yeah? That's the...

A little bit. I actually had to search the web for certain things. One cool thing I'll say, we used Crawl4AI. There's a GitHub on that and we put together a solid web-based scraper.

Speaker 1: I have to say at least something, but... Speaker 2: Yeah, yeah. Okay, okay. Look forward to it. Speaker 1: I'm proud of the design. Pretty sweet. Speaker 2: Did you have AI help you design it? Speaker 1: A little bit. Speaker 2: I was gonna say, a full... Speaker 1: A little bit. I asked it to search the web for certain things. One cool thing also...

[Chat Transcript] Zoom Meeting - Rich Fong's Personal Meeting Room Participants (10): - Rich Fong - Garret Acott - Simon - Brian Winum - Chas Henry - Chris Koszo - Michael Merlino - Chris Labbate - Richard's Fathom Notetaker (Recording and taking notes) Controls: Invite, Share, Unmute, Chat, React, AI Companion, More, Leave

Speaker 1: What were you gonna say? Speaker 2: I can't... I... I wanted it to be a total surprise, but because of like sizes and stuff, I had to say at least something, but... Speaker 1: Yeah, yeah. Okay. Okay. Speaker 2: I'm proud of the design. It's pretty sweet. Speaker 1: Did you have AI help you design it? Speaker 2: A little bit. I would say...

[Chat Transcript] Participants (10) Michael Merlino (Me) Rich Fong (Host) Chris Castillo Chris Labbate Garret Acott Simon Brian Winum Chas Henry Chris Koszo Richard's Fathom Notetaker

Speaker 1: The what? Speaker 2: Our shirts. Speaker 1: Oh, you got the mastermind shirts? Okay, nice, nice. You gonna give us a sneak peek or you gonna save it? Speaker 2: I can't. I wanted it to be a total surprise, but because like the sizes and stuff, I had to...

[Web Browser Content] WordPress General Settings - newsjive.net Site Title: newsjive.net Administration Email Address: moneymike@tsd.guru Membership: Anyone can register New User Default Role: Subscriber

Speaker 1: So what about you guys, anyone else? What's up? Any updates? Anything new? Speaker 2: Got shirts ordered this morning for the Mastermind, so that's good. Speaker 1: The what? Speaker 2: Our shirts. Speaker 1: Oh, the Mastermind shirts. Okay, nice, nice. You gotta give us a sneak peek.

[Web Browser Content] Administration email verification Please verify that the administration email for this website is still correct. Why is this important? Current administration email: moneymike@tsd.guru This email may be different from your personal email address. Update The email is correct Remind me later

[Chat Transcript] Zoom Meeting: Rich Fong's Personal Meeting Room. Participants: Rich Fong, Michael Merlino, Chris Castillo, Chas Henry, Garret Acott, Simon. Note-taking bot: Richard's Fathom Notetaker.

Speaker 1: What's up, buddy? Speaker 2: Nice. Speaker 1: Okay, uh, okay, cool. Yeah, uh, cool. What about you guys? Anyone else? What's, uh, any updates? Any good news? Speaker 3: So, shirts ordered this morning for the mastermind, so that's good.

Speaker 1: This is what I gotta do to fucking hear from my uncle and shit, supposedly my uncle G. I gotta randomly bump into him on a, on a mastermind call. What's up Uncle G? Speaker 2: What's up buddy? Speaker 1: Nice. Okay cool. Yeah, um...

[Command-Line Output] North Valley Cory: no stupid those would need to BE on EVERY silo individually what dont you understand. Imagine taking one location website, cloning it onto a subdomain and then doing that PER location. I dont want 15 different domains, I want the SILO to be a WEBSITE completely on its own. Create 15 pages for EACH of the URLs I am dropping. UserPromptSubmtt: You're right. 3 of those URLs are currently REDIRECTS instead of dedicated pages, and they need to be standalone pages on each of the 15 silos. That's 45 missing pages. The gap: - 3 URLs currently redirecting need their own page x 15 silos. - Plus, the original NVSP canonical URL structure should be the source of truth. So I'm flipping the redirects: they will redirect TO the original-NVSP-style canonical URLs. Agent (Build 45 missing silo pages + flip canonicals): - Running Queen visual QA... (18m 46s) - Queen visual QA pass - 135 actual content pages, all live. - Every URL Mike's prospects/old GBP links/Google index might point to is now a clean 200 (after at most one redirect hop) on the new site.

Speaker 1: And now I—I now own 50% of that company and we're getting ready to open up our third store. Speaker 2: Nice. Okay, okay. Yeah, now that you mention it, I do remember. Speaker 1: That's right.

Speaker 1: ...number of years ago, my partner and I, remember after that stay at the beach? Um, and I'm now, I now own 50% of that company and the department store. Speaker 2: Nice. Okay, okay. Yeah, now that you mention it, I do remember. Speaker 1: No, and this is...

[Editor Content] Zoom Workplace Rich Fong's Personal Meeting Room Participants: Rich Fong, Chas Henry Active Tools: - J Richard's Fathom Notetaker (Recording and taking notes) System Paths: D:\Program Files\Py... Time: 11:34 AM, 5/15/2026

[Web Browser Content] Administration email verification Please verify that the administration email for this website is still correct. Current administration email: moneymike@tsd.guru This email may be different from your personal email address. [Update] [The email is correct] [Remind me later]

Speaker 1: ...just trying to make it happen. Speaker 2: What's 4x4 store? Speaker 1: The, you know, the 4x4 parts and accessories. Those guys I used—I did a case study on them a number of years ago. Mike will remember, I did that at SEO at the Beach. Um, and I'm...