Loading video...

Video Failed to Load

Go Home

Cancel your influencer marketing platform 🤯 Claude Code can now scrape Instagram & TikTok for you Give it a niche and it goes out, finds the creators, pulls their real engagement metrics, digs up their contact info, and hands you back a live dashboard. All inside Claude Code. Perfect...

52,336 views • 1 month ago •via X (Twitter)

26 Comments

Kevin Wang's profile picture
Kevin Wang1 month ago

or you could connect to instead and do this for half the cost

Christopher Marcus's profile picture
Christopher Marcus1 month ago

has anyone tried this yet or it this just another gimmick

Sam's profile picture
Sam1 month ago

We've got a bunch of DTC tools and automations like this built into @typedotcom with free credits!

Tjay's profile picture
Tjay1 month ago

Can use apify for scarping they $500 in free credits, also works super well with Ghostfeed MCP if you want to create clone content

Reema Singhal's profile picture
Reema Singhal1 month ago

The real advantage is not scraping more creators, but replacing vanity metrics with comparable engagement and contact intelligence that improves who you actually approach.

Jake Morrison's profile picture
Jake Morrison14 days ago

The scraping side is solved. The real time-sink is the step after: which of those 200 creators will actually hit a brief on time and not ghost after the product ships. Would love to see someone crack vetting for reliability, not just reach.

Internet Labs's profile picture
Internet Labs1 month ago

Influencer marketing platforms charging $1,500/month watching a CLI terminal obliterate their entire SaaS business model with one free MCP connector.

Nick's profile picture
Nick1 month ago

real engagement rate beats follower count

Vai S. (AI Task Agents for Agencies)'s profile picture
Vai S. (AI Task Agents for Agencies)1 month ago

Wow! Does it also match the audience? If not, I’d transcribe the videos as well to shortlist influencers who align with the brand’s voice and mission.

🏝 JD Lasica's profile picture
🏝 JD Lasica1 month ago

Dude, this is crazy.

Mike Shi's profile picture
Mike Shi1 month ago

The interesting next step is making the connected tools legible: what they can access, what permissions they need, and whether the workflow is reproducible for another user. “Works once” is a very low bar for agent workflows.

Oliver Abas's profile picture
Oliver Abas1 month ago

informative

Saman Ahmed's profile picture
Saman Ahmed1 month ago

Follower count is such a weak signal. Actual engagement plus recent content gives you a much better shortlist.

Meadow Labs's profile picture
Meadow Labs1 month ago

Finding creators was never the bottleneck. Getting them to reply is. Though pulling real engagement at least stops you pitching someone whose followers are bought.

AI Mastery Guide's profile picture
AI Mastery Guide1 month ago

Days of manual sourcing gone, that's huge 🙌

daniel's profile picture
daniel1 month ago

influencer platforms discovering Claude has outbound now comment-mined hooks still need a place to go -

Write Wing Media's profile picture
Write Wing Media1 month ago

Code is Good. But don't forget to research, fact check, whet, and be the human in the loop. Don't automate everything. Then you will never stand out.

Adam Argaman's profile picture
Adam Argaman1 month ago

this is the real shift, agents doing the actual client work instead of just another dashboard

BeeReach's profile picture
BeeReach1 month ago

looks like a huge time saver for creator sourcing, especially when you need reliable engagement data across platforms

Salty's profile picture
Salty1 month ago

Will it work for X too? 👀

Adrian | The Web Scraping Guy's profile picture
Adrian | The Web Scraping Guy1 month ago

👀👀

Jadu 👐's profile picture
Jadu 👐1 month ago

same actor, new scenes — honestly the consistency is the whole cheat code lol

_alphashark_'s profile picture
_alphashark_1 month ago

Thousands of creator profiles pulled from one niche run makes me wonder what rate limiting ScrapeCreators hits before Instagram starts blocking the requests.

Sogni.ai's profile picture
Sogni.ai1 month ago

finding the creators is half the job, the other half is having creative ready when they say yes 😄 that part's solved too, Sogni Creative Agent renders the assets inside the same claude setup: npx setup-sogni-agent-skill

George Rowan's profile picture
George Rowan1 month ago

Seems illegal at first

Adrián's profile picture
Adrián1 month ago

Quiero

Related Videos

I just built a Claude Code plugin that automatically researches viral trends in any niche 🤯 Drop in your niche -> and it surfaces the hooks, angles, and topics going viral right now across Reddit, YouTube, TikTok & Reels, so you always know exactly what to post. All inside Claude Code. Perfect for DTC brands, agencies, and creators who are tired of guessing what to make. If you're spending hours "researching" content, doom-scrolling TikTok for ideas, copying competitors who are also just guessing, staring at an empty content calendar every Monday... This plugin kills the entire loop: → Drop in a keyword (your product, your niche, a competitor) → It scrapes Reddit, YouTube, TikTok & Instagram Reels → Ranks every post by real engagement → Pulls the hooks, pain points, & exact customer language that's working → Renders a clean content-brief dashboard telling you what to make No guessing. No doom-scrolling. No blank calendar. I ran it live on "magnesium for sleep" — in 3 minutes it surfaced a 5.1M-view hook and the exact complaint customers keep repeating ("it works for a few weeks, then stops"). What you get: → The top hooks already winning, with real view counts → Your customers' pains in their own words → The formats getting watched right now → 3 ready-to-shoot content ideas Built 100% in Claude Code (with Apify + Firecrawl). I created a step-by-step playbook that shows you how to build it yourself. Want access for free? > Like this post > Comment "TRENDS" And I'll send it over (must be following so I can DM)

Mike Futia

79,548 views • 4 months ago

I just vibe-coded a TikTok slideshow automation app in Claude Code 🤯 One-click generation of the exact slideshow format that's getting DTC brands and agencies millions of organic TikTok views right now. Niche prompt + Pinterest images + Claude = finished slideshows, ready to post. Yep, I basically cloned Reelfarm in a weekend with Claude Code. Perfect for DTC brands and agencies who want to farm organic TikTok reach without hiring a creator or filming a single video. If you've watched accounts blow up posting nothing but static images with text overlays — hundreds of thousands of views per post, no face, no camera, no editing — and wondered how the hell they're producing that volume... This app eliminates the entire loop: → Build an "automation" around any niche (finance, beauty, fitness, relationships) → Claude writes the hooks, body slides, titles, and captions in your voice → Scrape Creators pulls on-aesthetic images from Pinterest → App stitches slides together with text overlays, hook image, and CTA → Auto-publishes to TikTok via Blotato (or exports as drafts) No creators. No filming. No editing in CapCut. What you get: → A reusable content engine — set the niche once, get endless slideshows → A hook bank that auto-refills when it runs low → Image collections sourced from Pinterest (not AI-generated slop) → Built-in editor to swap images or tweak copy before posting → Optional scheduler that posts for you on autopilot Built 100% in Claude Code. Deploys to Replit in one click so it runs 24/7. I put together a starter template pack with all of the prompts so you can vibe-code this yourself Want the prompts for free? > Like this post > Comment "SLIDES" And I'll send it over (must be following so I can DM)

Mike Futia

43,354 views • 5 months ago

I built an Agency Client Finder in Claude Code that finds DTC brands who need a creative agency, and writes the pitch for you 🤯 Drop in a niche → it scans the Meta Ad Library, finds brands that are spending on ads but running weak creative, and hands you a ranked prospect list with a tailored pitch for each. All inside Claude Code. Perfect for agencies who are great at the work but terrible at the pipeline. The hardest part of running an agency was never the creative, it's finding clients who actually need you AND have the budget to pay. So you cold-DM brands that ignore you, scrape LinkedIn for "Head of Growth" titles, and pitch brands that already have an agency or no budget at all. This finds the ones already raising their hand: → Drop in your niche (skincare, supplements, coffee, pet, men's grooming…) → It scans every brand running ads in the Meta Ad Library → Scores each one: spending on ads, but weak creative — low volume, stale, catalog-only, no video → Pulls their actual ad creatives and critiques them like a creative director → Looks up each brand's email + website so you can actually reach them → Hands you a ranked list with the evidence, a pitch, and the contact info No cold-list spray-and-pray. No pitching brands with no budget. No guessing who actually needs you. What you get: → A ranked list of brands spending on ads but starved for creative → Their real ads, with the exact creative gap named for each → A tailored pitch angle per brand — walk into the call with the evidence → Their actual email + website, so it's an outreach list, not a research list → Run it on any niche, any market, in minutes Built 100% in Claude Code, on the Meta Ad Library. I put together a step-by-step playbook showing you how to build this: the Ad Library scrape, the scoring, the whole pipeline. Want the playbook for free? > Like this post > Comment "CLIENTS" And I'll send it over (must be following so I can DM)

Mike Futia

23,086 views • 3 months ago

THIS MIGHT BE THE #1 OPEN-SOURCE REPO FOR CLAUDE CODE RIGHT NOW. IT GIVES CLAUDE A MEMORY AND SLASHES YOUR TOKEN COST ON EVERY QUESTION The repo is safishamsi/graphify, a free open-source skill that turns any codebase into a knowledge graph Claude Code can read instantly. Instead of grepping through your files every session, Claude gets a map of how everything connects The problem it fixes: Every time you ask Claude Code about a big repo, it does the same thing, greps through dozens of files like a brute-force Ctrl+F, blows through your context window, and sometimes still misses the answer hiding in a file nobody searched. Claude Code has no memory of how your project is structured. Every session starts from zero What it does: It maps your entire codebase into a knowledge graph, capturing not just which files exist, but which functions depend on which, which modules are central, and which files cluster around the same concern. Claude queries the map instead of scanning files How it works, three passes: 1. Code structure, free and local. Tree-sitter parses your files and pulls out classes, functions, imports and call graphs. No LLM, no tokens, just your actual code mapped deterministically 2. Audio and video, if you have them. Transcribed locally and folded into the graph 3. Docs, papers, images. Here an LLM does semantic analysis, figuring out what each document means and where it fits. Only the meaning gets sent up, never your raw source It saves you money: Normally a question about a big repo makes Claude spawn explore agents that scan file after file, eating your context window and your token budget before you get an answer. With the graph already built, Claude queries the map instead of re-reading the codebase every time. Same answer, a fraction of the tokens. The graph only gets built once, then a hook rebuilds it after each commit for free, so you never pay that scanning cost again. The bigger the repo, the bigger the gap The best parts: it's a skill, so once installed Claude knows when to use it without you memorizing commands. It works on non-code folders too, point it at docs or notes and it can spin up an Obsidian vault How to add it to your Claude: 1. Install Claude Code if you haven't: npm install -g Paul Jankura-ai/claude-code 2. Add the skill: claude skill add safishamsi/graphify 3. Open your project folder and run /graphify . to build the graph 4. Optional, make it automatic: graphify hook install so the graph rebuilds after every commit That's it. Ask Claude about your repo and it reads the map instead of burning tokens on a file hunt Bookmark this

Yarchi

56,502 views • 4 months ago

I just built an AI-powered creative search engine with Gemini Embedding 2 + Claude Code 🤯 Drop in your UGC clips, product shots, and ad variations — then search through everything in plain English. "Show me all the unboxing clips." "Find product shots with natural lighting." "Which creator talked about sensitive skin?" All inside Claude Code. Perfect for DTC brands and agencies sitting on hundreds of creative files they can never find when they actually need them. If you're digging through a folder of random file names, scrubbing through raw footage to find that one clip, and relying on memory to track down what's already been shot... This system eliminates the entire loop: → Drop your videos, images, and docs into a project folder → One prompt to Claude Code — it builds the entire search app for you → Google's new Gemini Embedding 2 model actually watches your videos and looks at your images → It understands what's inside each file — not just the file name → Search in plain English and get back the actual assets with confidence scores No scrolling through folders. No relying on file names to find anything. No re-shooting footage you already have. What you get: → A searchable library of every creative asset your brand has ever produced → Natural language search across video, images, and documents at the same time → Results that show the actual files inline — play videos, view images, read docs → A system that gets smarter every time you add richer descriptions to your assets One free API key. No monthly subscriptions. Runs on your machine. I put together a full playbook with the exact build prompt, the setup process, and DTC/agency use cases to get this running in under 30 minutes. Want the full playbook? > Like this post > Comment "SEARCH" And I'll send it over (must be following so I can DM)

Mike Futia

12,305 views • 7 months ago

I just built a Viral Radar Tracker in Claude Code that finds overperforming TikToks, Reels & Shorts 🤯 Track any creator, hashtag, keyword, or niche → it pulls every video into one research feed and ranks them by outlier performance: how hard each video overperformed vs. that creator's normal baseline. All inside Claude Code. Perfect for DTC brands and creative strategists who need a steady pipeline of proven hooks and angles for ads and UGC. If your "creative research" is sorting TikTok by view count, seeing the same 10 mega-creators every time, screenshotting videos into a Slack channel nobody opens, then walking into the creative meeting with a blank page anyway... This radar replaces the entire loop: → Track creators, hashtags, keywords & niches across TikTok, Reels & Shorts → Every new video lands in one research feed → Ranked by outlier score, not follower count → Save the winners into swipe files grouped by theme → Break down WHY each one worked: hook, angle, format, proof No sorting by raw views. No chasing already-huge accounts. No blank-page strategy sessions. What you get: → The videos overperforming in your niche right now → The hooks and formats behind them → A living swipe file organized by theme → Winning patterns turned into ad concepts and UGC briefs Built 100% in Claude Code. I put together a complete "prompt pack" with 25+ prompts so you can build this yourself + a walkthrough video. Want access for free? > Like this post > Comment "VIRAL" And I'll send it over (must be following so I can DM)

Mike Futia

10,379 views • 3 months ago

I just built a Reddit research agent in Claude Code that turns real customer complaints into ad angles 🤯 It scrapes the threads where people actually rant about your category, reads every post and comment, and hands you back a dashboard of pain points, exact customer phrases, and ready-to-test ad angles. All inside Claude Code. Perfect for DTC brands and agencies who want ad angles pulled from real customer language, not guessed in a vacuum. Reddit is the most honest focus group on earth. People say things there they'd never put in a survey or a review with a brand watching. So if you're writing hooks off your own assumptions — or digging through Reddit by hand, opening tab after tab, pasting quotes into a doc... This agent runs the entire loop: → Give it a product or category + a couple of subreddits → It searches Reddit for the threads where people actually discuss the problem → Pulls every post and comment → Claude reads all of it and finds the patterns → Builds a dashboard: pain points, desires, objections, and swipe-worthy phrases No manual scrolling. No pasting quotes into docs. No guessing what your customers actually care about. What you get: → Ranked customer pain points, each with a real verbatim quote → The exact language your customers use, ready to drop into ad copy → Objections and buying triggers pulled straight from the threads → 8-10 ready-to-test ad angles built from all of it Built 100% in Claude Code. Want the skill file for free? > Like this post > Comment "REDDIT" And I'll send it over (must be following so I can DM)

Mike Futia

46,275 views • 3 months ago

Claude Code + computer use is f*cking cracked 🤯 Build a landing page → Claude opens Chrome, looks at it, spots every issue, and fixes it — without you describing a single thing. All inside Claude Code. Perfect for DTC brands and agencies who are still vibe-coding landing pages and advertorials in Claude Code, then manually opening them in Chrome, spotting 15 things wrong, and describing every visual issue back to Claude one at a time. If you're building pages in Claude Code and your workflow looks like this — build the page, open it in Chrome, spot broken spacing, go back to Claude, type "the CTA button is too low and the hero image is cut off," wait for the fix, open Chrome again, find 3 new issues, describe those too ... Claude Code + computer use eliminates the entire loop: → Claude writes the full landing page or advertorial → Opens Chrome and navigates to it → Spots layout issues, broken spacing, off-brand colors, missing elements → Fixes everything and re-checks until the page looks right → Tests your Shopify product pages by clicking through like a real customer → Walks through your checkout flow and flags friction before customers hit it → You only see the finished, visually verified result No describing what you see on screen. No "the CTA button needs more contrast" back-and-forth. No being the eyeballs for an AI that can't see. What you get: → Landing pages and advertorials Claude builds AND visually QAs before you ever look at them → Product pages Claude clicks through — testing layout, images, and CTAs like a real user → HTML dashboards Claude opens and verifies the charts actually render → Checkout flows Claude walks through step by step to catch friction → All of it happening in one session — build, test, fix, done One prompt. Claude builds it, checks it, and fixes it. You just review the finished page. I put together a full playbook with the exact setup, the prompts, and 5 DTC workflows that use Claude Code + computer use. Want it for free? > Like this post > Comment "CLAUDE" And I'll send it over (must be following so I can DM)

Mike Futia

19,249 views • 6 months ago