maxppap

@claudeai well finally

New in Claude Code: agent view. One list of all your sessions, available today as a research preview.

working on this exact problem from the surface angle — AI that lives inside the user's existing CRM as a browser extension instead of in its own chat window. when AI shares your DOM, output format flattens (it edits what you're looking at) and gesture comes for free. the progression i'd extend yours with: text → markdown → HTML → same surface. format matters less than where the output lives.

This works really well btw, at the end of your query ask your LLM to "structure your response as HTML", then view the generated file in your browser. I've also had some success asking the LLM to present its output as slideshows, etc. More generally, imo audio is the human-preferred input to AIs but vision (images/animations/video) is the preferred output from them. Around a ~third of our brains are a massively parallel processor dedicated to vision, it is the 10-lane superhighway of information into brain. As AI improves, I think we'll see a progression that takes advantage: 1) raw text (hard/effortful to read) 2) markdown (bold, italic, headings, tables, a bit easier on the eyes) <-- current default 3) HTML (still procedural with underlying code, but a lot more flexibility on the graphics, layout, even interactivity) <-- early but forming new good default ...4,5,6,... n) interactive neural videos/simulations Imo the extrapolation (though the technology doesn't exist just yet) ends in some kind of interactive videos generated directly by a diffusion neural net. Many open questions as to how exact/procedural "Software 1.0" artifacts (e.g. interactive simulations) may be woven together with neural artifacts (diffusion grids), but generally something in the direction of the recently viral x.com/zan2434/status… There are also improvements necessary and pending at the input. Audio nor text nor video alone are not enough, e.g. I feel a need to point/gesture to things on the screen, similar to all the things you would do with a person physically next to you and your computer screen. TLDR The input/output mind meld between humans and AIs is ongoing and there is a lot of work to do and significant progress to be made, way before jumping all the way into neuralink-esque BCIs and all that. For what's worth exploring at the current stage, hot tip try ask for HTML.

@techbromemes before, you just googled it. the whole paradigm of search and working with data has changed

@remain_urus good — better setup imo: bundle the whole harness + agents into one MCP server for ollama models. claude code becomes the orchestrator, ollama-via-MCP is the worker pool

If you're on a tight budget and don't want to spend 100-200$ on codex/claude, try ollama.com/pricing You can use the 20$ sub in most tools + openclaw/hermes, and the limits you get for that money are insane, you get the newest kimi, glm, deepseek models and more

@1Umairshaikh marketing — but more specifically, naming a pain people already feel. AI products that vaguely "help" with stuff go nowhere.

The hardest part of AI startups? Not building the product. Convincing people it’s worth paying for.

@EXM7777 most people haven’t even typed their first prompt into chatgpt yet

i can't believe a year ago we were still copy-pasting prompts inside chatgpt

@0xPrajwal_ btw u already can use Claude for everything from this list

✅ Claude Code ✅ Claude Cowork ✅ Claude Design ◻️ Claude Marketing ◻️ Claude Sales ◻️ Claude HR ◻️ Claude Finance ◻️ Claude Customer ◻️ Claude GF

@asaio87 because someone needs to own the system. when cloudflare dies, a $10/hr freelancer can’t fix it — only someone who’s been inside every layer for years can. the $360k isn’t for the code, it’s for the speed and accountability when things break.

Vibecoding is so easy with AI, and you can build apps so fast right now. Correct? Then why are companies still hiring developers for $360k/year and not hiring any freelancers who work for $10/hr ?

@yashhq_22 learn to lay a roof

In 2020: learn to code. In 2023: learn to prompt. In 2026: learn to build with agents. In 2029: ???

@OpenAIDevs voice control is one half of CRM-as-AI-surface. building the other half — analytics + insights, browser extension, no API keys. these two together would be unreal eventually wakelead.com

Here’s how you can integrate GPT-Realtime-2 to bring voice control to a CRM workflow.

CRMs store everything. They tell you nothing. I'm building the analytics layer they're missing — a browser extension that lives in your CRM sidebar, sees your data directly (no API keys, no setup), and turns it into actual insights. Now: → creates deals, contacts, tasks from one sentence → instant analytics on existing CRM data → insight discovery beyond the funnel view → fully client-side, data never leaves your browser Next: deep research mode — bottlenecks, growth points, churn risk, payment audits, stale deals, 10-step quarterly audits with action items. HubSpot, Pipedrive, Salesforce, Google Sheets, custom skills. Pipelines show stages. I want to show you why deals stall, who's about to churn, and where your next 20% of revenue is hiding. Short demo below ↓ This is the idea I'm most excited about in years, and I'd rather build it with you than guess. Tell me: — what's the one CRM workflow you'd hand off to an AI tomorrow? — what process do you wish you'd never have to touch again? — what insight do you keep wishing your data could surface on its own? Reply, DM, whatever's easier — I read every message. Looking for early users + design partners who want a real say in what ships. wakelead.com

@trikcode broke the loop. shipped this week hunting for feedback wakelead.com

the vibe coding cycle in 2026 Day 1: build the app. Day 2: buy the domain. Day 3: design the logo. Day 4: notice 200 competitors. Day 5: start over with a new idea. Repeat. Ship nothing.

@heyblake wakelead.com

Drop your project URL Let’s drive some traffic

@LoopandPixels AI CRM assistant, browser extension no API keys, no setup, no MCP wakelead.com

What are you building currently? - Website - Android app - IOS app - Nothing

@nxhaaa19 @X just shipped Wakelead — AI CRM assistant as a browser extension. no API keys, no setup, surfaces insights beyond the funnel view. would love any feedback 🤝 wakelead.com

Solo builders where you at 🙋‍♂️ @X Looking to connect with people who are... Shipping without permission: 💻 SaaS products 🤖 AI wrappers ⚡ Automation tools 📱 Mobile apps 🛠️ Dev tools No VC. No co-founder. No excuses. Drop your project 👇 Let's grow together.

@sflorimm the team. you can clone a product in a weekend. you can't clone the people who keep pushing when it's hard.

What makes a startup uncopyable?

great breakdown for the agency play. went a different route with my own product — browser extension inside the CRM directly, so no Zapier/Make/n8n layer in between. data stays client-side, install in 30 sec. tradeoff is i'm locked to one shape (CRM insights) but setup friction drops to zero. just shipped, any feedback welcome 🙏 wakelead.com

The 3 ways to build a speed-to-lead agent in under a day: 1. Zapier. 1-2 hours to build. Best for simple SMS-to-CRM flows for non-technical clients. Gets pricey at scale ($50-150/mo). 2. Make. 2-4 hours. Native AI modules, visual builder, $20-30/mo covers most use cases. The sweet spot. 3. n8n. 3-5 hours. Self-hosted option, no per-execution fees, full conversational agent capability. Pick Make if you're building for paying clients. Best balance of power and price. Full breakdown in the article below.

@DavidOndrej1 thx man! funny timing — just shipped my first public build today. AI CRM assistant that helps find insights, runs as a browser extension. open to feedback or advice from literally anyone 👀 wakelead.com

stop developing locally start developing on a VPS trust me

@bcherny @kaiofreitas I’ll go to try it before sleep

@kaiofreitas /loop until

hey @bcherny when /goal for CC? 🫣

@sama a model that tells me my idea is bad without first explaining how brave i am for asking

what would you most like to see improve in our next model?