ElWalka

OpenClaw: free Mac Mini: $599 Claude Max: $200/mo Keyboard: $25 Monitor: $195 Mouse: $20 having to rebuild your entire agentic stack on a holiday weekend because Anthropic moved the goalposts: priceless

Governor Abbott drained an entire billion dollars from the state's public school system for this. 😒

Anthropic is shutting down subscription usage for tools like OpenClaw tomorrow. Fortunately I can swap in Codex in my config and all my bot’s skills still work as expected. Build your stacks to avoid vendor lock-in.

🚨BREAKING: Block (Jack Dorsey's company) just open-sourced a local AI agent that goes way beyond code suggestions. It's called Goose and it installs, executes, edits, and tests with any LLM fully on your machine. 100% Opensource.

Drop everything you're doing and watch this now. RIGHT NOW. Ari K and Schuyler Brown, I have no idea who you 2 beautiful bastards are, but you're instant legends. Hopefully someone in the comments knows and tags.

🚨Breaking: Accountants charge $150/hour to sort receipts you photographed on your phone. TaxHacker does it with an LLM -- self-hosted, on your own server, with your financial data never leaving your control. It's called TaxHacker. And it's not a budgeting app. It's a self-hosted AI accounting tool that takes photos of receipts, invoices, and bank statements -- any language, any currency, even handwritten -- and extracts every field you need into a structured database automatically. Here's what it actually does: → Upload a receipt photo or invoice PDF -- AI extracts date, amount, vendor, line items, tax, and saves it all → Auto-categorizes every transaction based on content → Converts 170+ currencies and 14 cryptocurrencies (BTC, ETH, LTC, DOT) using historical rates from the actual transaction date → Custom fields with your own AI prompts -- add any column you need, extracted automatically → Full-text search through recognized document content → CSV export with attached documents for your accountant → Works with OpenAI, Google Gemini, or Mistral Here's the wildest part: Every LLM prompt is editable. The system prompts, the field extraction rules, the category logic -- all of it is exposed in settings and fully customizable to your industry, your country, your specific document types. One command to self-host: `docker compose up` 1.1K GitHub stars. 10 releases. Built for freelancers and indie hackers. 100% Open Source. MIT License. (Link in the comments)

This guy literally automated his entire social media with one free AI agent. No team. No scheduling tools. No VA. Here's exactly how he did it 👇 The agent is called Hermes. It's free and open source. Here's what it's running on autopilot right now: → Scans Reddit every 24 hours for trending topics → Writes posts based on breaking news → Generates custom images with Nano Banana 2 → Posts directly to X, TikTok, and Instagram with captions and hashtags → Pulls daily analytics and flags what's working → Researches competitors and shows what's getting engagement One post already hit 1,000 views. He didn't write it. He didn't post it. He didn't even check it. Here's the part most people skip: Day one it won't be good. You give it feedback. It writes notes on what went wrong. It improves the skill file. By day 7 it's better. By day 30 it knows your voice. It's called a closed learning loop. Every task makes the next one sharper. The setup takes an afternoon: Install Hermes from GitHub. Connect a free Qwen 3.6 API key from OpenRouter. Add Firecrawl for web research. Create your first skill. Give it feedback. Let it run. That's the whole system. Free. Open source. Gets better every single day. Save this. Then go install it.

Do you even understand what this means? An open source model just released that is: • Outperforms models 20x its size • Can run on a base model Mac Mini • Is AMERICAN 🇺🇸 If you have a base model Mac Mini you can have unlimited super intelligence on your desk. For free. Sonnet 4.5 was released 5 months ago In 5 months that level of intelligence went from frontier to free on your desk And not only that, can run on any basically any computer out there If you have even a remotely modern computer, do the following immediately: 1. Download LM Studio 2. Go to your OpenClaw and ask which of these new Gemma 4 models is best for your hardware 3. Have it walk you through downloading and loading it 4. Build apps with it knowing you are using your own personal, private super intelligence on your desk The people denying this is the future are so beyond lost.

🚨BREAKING: The team behind a marketing agency generating millions in revenue just open sourced their entire Claude Code skill set for growth experiments, sales pipeline, content ops, outbound, SEO, and finance automation. It's called AI Marketing Skills and these are not prompts or toy demos. They are complete Python workflows with real statistical methods, scoring algorithms, expert panels, and automation pipelines that you drop into Claude Code and run against live business operations today. The Growth Engine skill uses bootstrap confidence intervals and Mann-Whitney U tests on marketing experiment results, which is actual statistics, not vibes. The Deal Resurrector tracks contacts who leave companies and routes outreach to their new employers automatically. → Expert Panel skill recursively scores content with domain-specific personas until quality hits 90 or above → ICP Learner rewrites your ideal customer profile automatically based on actual win and loss data → RB2B Router does intent scoring, seniority dedup, and agency classification before routing to outbound sequences MIT License. 100% Opensource. Link in comments.

We built a hedge fund entirely on AI.

Here's an update so far on the automated research pods! Emergency update! So unfortunately, and hilariously I discovered that the in-warehouse ambient CO2 levels are 700! What does this mean for the pilot trial? Well, since pod one was supposed static enrichment - at 700 ppm of CO2 injected that meant that no supplemental injection was needed. And since pod 4 was supposed to be the control - or ambient(420ppm) we got 700! Basically - pod 1, 2 and 4 got *almost* identical CO2 treatments. So 3 controls, and one treatment... Oops! BUT here's the interesting thing. Pod 3. We did validate something with this brief trial from seedling to vegetative. Pod 3 had 1000 ppm CO2 injected, with a par/light level of 100. And STILL has the same amount of growth as the other 3 pods running at basically full hog. That's 57% LESS light energy than the other pods. That's huge cost savings already, where you supplement long light hours in place of CO2. This validates the research done by NC state, Huber et al. 2021, Frontiers in Plant Science (link below) To put it very simply - HUGE cost savings for indoor growing, just by supplementing CO2 in place of light. So what's next? I already have seedlings ready to go, and we'll transfer them for another pilot study - it will be tailored to match the indoor environment of the warehouse itself. These tomatoes will be transferred to the enclosed grow room where they can begin to grow as productive tomatoes. Which is perfect since the leader/follower arm is coming soon, and we'll need plants in there ready to train the act policy. This also lets me work out some kinks in the hardware. Dying serial connections on the Arduino, and some other odds and ends. I'm also excited to share the new upgrade - which will be the camera mounted on a linear actuator. This allows the agent to determine the exact height of the plant, without the parallax effect leading to better observations from the agents, and lighting control that's reactive to the height of the plant. More soon! 🍅🤖🧪🤩

Six months ago, I had an app idea but no developer, no budget, and no clue where to start. Today, I built the whole thing for free in under an hour. Here's what changed. I discovered Google Stitch and Google Anti-Gravity. Stitch is an AI-powered UI design tool from Google Labs. You describe your app, and it generates real structured UI not a flat image, actual components with colors, spacing, and interactive states. Anti-Gravity is a standalone AI coding environment. It doesn't just suggest code. It dispatches agents that plan, write, test, and verify your app autonomously. The moment Stitch exports your front end directly into Anti-Gravity via MCP protocol that's when it clicks. The design and the code are talking to each other. And you're just in the manager seat. This isn't the future. It's available right now. And it's free. Save this video, you'll need it the day you finally decide to build. 🔖 Want the SOP? DM me.

the AI version of what Kevin is describing: 1. pick a niche (dentists, gyms, med spas, roofers) 2. use Claude to research their competitors, audience, and top-performing content 3. build a content system that creates 30 days of short-form ads in one afternoon 4. run the ads. track customer acquisition weekly. show them the numbers. 5. charge $3-5K/mo per client because you're not selling "content," you're selling customers Kevin says these people make $250-500K/yr doing this manually. with AI handling the research, scripting, and repurposing, you can run 5-10 clients at once instead of 2-3. same skill. 3x the capacity. no team needed.