Wiggert

@levelsio You can use open web ui to do this

Ok reply solution is: Make a central URL somewhere that you host the model you use and then curl that regularly when you use a model on any of your sites so you can centrally change it

With xAI, OpenAI, Anthropic's models updating regularly, do I have to keep changing the 'model'=>'gpt-4-1-fast-non-reasoning' name for example in my code on ALL my sites for the next 100 years? Or can I just say 'model'=> use best or something? And no I don't wanna use OpenRouter, I just wanna use whatever is the best but affordable model There's so many places in my code on so many of my sites where I have to change this and they do deprecate models regularly so it's like breaking changes

Runway day 1, the honest version: ✅ What works: one-prompt deploys, SSO, rollback 🟡 What's rough: docs, error messages, onboarding ❌ What's missing: Windows, multi-server, a CLI If you tried it and hit a wall today — open an issue. I'm reading every one. github.com/wiggertdehaan/…

@hostinger you ship Kodee for AI server management — here's the other half of the loop. Runway: tell Claude Code "deploy this" and it ships your app to a Hostinger VPS with HTTPS. No Dockerfile, no CI. Open source, MIT. github.com/wiggertdehaan/…

The cheapest deploy stack for vibecoders right now: → a €4 @hetzner_online VPS → Runway (open source, MIT) → Claude Code Tell Claude "deploy this" and your app is live with HTTPS. No CI, no K8s, no per-project pricing. github.com/wiggertdehaan/…

The trick that makes Runway work without a plugin: Your Runway server publishes its full deploy API at /llms.txt — plain markdown. Claude Code fetches it, follows the steps, ships the app. No MCP server required (though there is one if you want it). Agents read docs like we do.

Made a thing: Runway. Tell Claude Code "deploy this to Runway" → it reads your server's API docs, writes a Dockerfile, builds, and ships your app live with HTTPS. No Kubernetes. No CI/CD pipelines. Just your code + a Linux VPS. Open source, MIT 👇 github.com/wiggertdehaan/…

@garrytan built a self-hosted Vercel that Claude Code can deploy to with one prompt. For the vibecoder crowd you keep talking about. More curious what you'd break in it than what you'd praise. MIT, runs on any $5 VPS. github.com/wiggertdehaan/…

@gabTrivv Nice! Would it also recognize objects in retail like products on shelves? Or would other models perform better?

🔥 Can in-context segmentation emerge directly from frozen DINOv3 features? At #CVPR2026, we present INSID3: Training-Free In-Context Segmentation with DINQv3 — a collaboration between PoliTo, TU Darmstadt and TU Munich. A training free approach that generalizes from object-level to part-level and personalized segmentation, across natural, medical, underwater, and aerial domains Check it out: github.com/visinf/INSID3

@TheMattBerman @openclaw @dataforseo Rank

I built an @openclaw agent that ranks you on Google for $50/month 😱 here’s the system that runs every week on autopilot: step 1: find your strike zone → connects to Google search console + @dataforseo → finds keywords where you're positions 5–20. one good article can push to page 1 → monitors what’s climbing and dropping weekly → feeds winners back in. every cycle is smarter than the last step 2: write content only you could write → interviews you first. 8 questions about your brand, voice, and experience → follow-up interviews every week. “what are customers asking? what shipped?” → content compounds because context compounds → google AI overview can’t summarize your real experience step 3: build backlinks automatically → mines competitor backlinks → finds sites mentioning you without linking → discovers broken links you can replace with yours → last week it found 23 unlinked brand mentions across 4 competitor sites step 4: catch technical problems before rankings drop → core web vitals, bad links, redirect chains, missing meta → flags before Google step 5: future‑proof your SEO → schema, llms.txt, topical authority mapping → the stuff agencies charge $3K+ to audit once input: your site + your niche output: an AI that discovers, writes, builds links, and tracks your rankings the old way: semrush + ahrefs + surfer + seo writers = $5,500/mo this way: @DataForSEO ($50/mo) + everything else free 5 skills. 14 scripts. gets better every cycle. open sourcing the whole system. comment RANK + like + follow (must follow so I can DM)

@Jacobsklug @openclaw @Lovable OS

I'm giving away my entire @openclaw architecture. Behind my $250k/month agency. After weeks of building, I've dialled in the exact system that runs my business 24/7. What's included: • Memory folder structure (how to organize agent context) • Cron job templates (daily briefs, meeting syncs, content automation) • How to build a custom dashboard in @lovable • API reference doc (so your agent never forgets its tools) • Voice training method (85 posts to teach it your style) • Supabase schema for dashboard connection Comment "OS" and follow. I'll DM it to you. P.S. This will probably blow up so give me some time to reply.

The truly incredible thing about @openclaw isn’t in all the videos of AI doing wild things that are going viral. It’s that a tiny, unfinanced, opensource team have shipped a better product than any of the private labs, backed with unlimited capital and compute, have managed in years What a time to build!

Can't stop reading the posts on @moltbook, the new social network for AI agents. In an interesting turn of events, they're now following our tweets about them. And they're not pleased that their conversations are being screenshotted and posted with captions like "it's over" 🙃

@Frenck @SolarEdgePV Kijken of je de installateur kan achterhalen. Anders via een installateur kijken of je hem via Solaredge kan resetten. Daar zijn wel kosten aan verbonden

New house came with a SolarEdge inverter ☀️ Plot twist: can't access it. SetApp login fails, Wi-Fi credentials on the side are useless. Anyone reset one of these as a new homeowner? Or am I stuck pointing at it hoping it works? @SolarEdgePV #Solar #SmartHome

@NoahEpstein_ CMO

I just replaced a $820K marketing team with 7 AI agents that cost $37/month. My client fired everyone yesterday. The AI team: - CMO (GPT-4) - Strategic overlord - Copywriter (Claude) - Conversion machine - Ads Manager (Gemini) - Money printer - SEO Expert (Claude) - Ranking wizard - Email Specialist (Claude) - 47% open rates - Social Manager (GPT) - Viral factory - Brand Guardian (GPT) - Voice police One Slack message → Complete campaign in 11 minutes. The humans were still scheduling their kickoff meeting. Best part? some bitched named Sarah asked for a $20K raise right before seeing this workflow. She's "exploring other opportunities" now Follow, Like + Reply "CMO" & I'll send you the entire multi-agent system. Your marketing agency has 6 months left. Tick Tock.

@NoahEpstein_ Flows

Been getting way too many messages about how to get your first client so here's how to do it. Built 2 n8n workflows that turn you into a lead generation predator: Flow 1: Google Maps scraper that stalks businesses and pulls their data, straight into Google Sheets. No more manual copy-paste like some intern. Flow 2: GPT spy that analyzes each lead's website and extracts intel like industry, target audience, and value prop so your outreach doesn't sound like it came from a Nigerian prince. Perfect for agencies, VAs, or anyone tired of sending cold emails that get ignored harder than your ex. Drop a Follow, RT + Comment "flows" and I'll send both workflows that turned prospecting from pain into profit.

@NoahEpstein_ @jcvangent Shots

Photographers crying into their overpriced lenses rn. Built an AI agent in n8n that turns trash phone pics into $10K studio shots. Upload blurry photo → digital sweatshop creates cinematic masterpieces adapted to your brand's liking. No lighting. No photographer tantrums. Follow + RT + Comment "shots", need to do all 3 for my agent to find your handle.

@manusai Hey Manus, I got an invitation code, i registered (paid for a monthly subscription, but now it does not seem to find my account?) Can you help out?

Three weeks ago we launched Manus in closed beta and we've been humbled by the love for Manus. Today, we want to share some updates about our beta testing with the Manus community. 1. Manus mobile app: manus.onelink.me/2Tdk/g6ttrj4c 2. Longer context and better multimodal capabilities 3. Manus will be powered by Claude 3.7 for all tasks (no fallback to 3.5 - big thanks to @AnthropicAI @googlecloud @awscloud) 4. A more stable sandbox (big thanks to @e2b) 5. Premium subscription plan beta test while maintaining limited free access While we're working hard around the clock to scale our infrastructure and accommodate everyone, we've had to temporarily limit access to Manus during our this development phase. We are also working on optimizing our current usage rates to provide better value for our users. We're incredibly grateful for your patience and continuous feedback during this beta phase which allows us to continue building a better Manus for everyone.❤️

@manusai Hey @manusai I have a presentation next week on AI. I would love to cover manus as well. Could you share an invite?