What it does

Twitter Personality is a web app that takes a Twitter handle, scrapes the account’s content through multiple fallback services (Twitter API, Apify, SocialData), and pipes it into Wordware AI prompts. The output is a generated personality profile — with variants including a “roast” mode and a “pair” comparison feature locked behind a password. Stripe handles payments, PostHog tracks events, Loops manages newsletters, and Neon stores the data.

The interesting bit

The scraping layer is deliberately over-engineered: three different services with fallback logic, plus Twitter cookie-based access as a last resort. That’s either resilience or an admission that scraping Twitter is a brittle business — the README doesn’t say which. The actual “AI agent” is mostly prompt orchestration via Wordware’s API, which is refreshingly honest as architecture goes.

Key highlights

• Multiple scraping backends (Twitter API, Apify, SocialData) with automatic fallback
• Several Wordware prompt modes: standard profile, roast, full analysis, and pair comparison
• Full SaaS plumbing: Stripe payments, PostHog analytics, Loops email, Neon database
• Next.js/TypeScript stack with a 1,448-star repo that appears to be primarily Wordware’s own marketing vehicle

Caveats

• The README is essentially a setup guide and env var checklist; no discussion of accuracy, privacy, or rate limiting
• “Cutting-edge AI technologies” is the project’s own phrasing — the actual mechanism is API calls to managed prompts
• Heavy external dependency footprint: running locally requires ~15 different API keys and services

Verdict

Worth a look if you’re building a prompt-wrapped SaaS and want to see how the plumbing fits together. Skip it if you were hoping for novel ML research or a scraper you can easily run without signing up for half a dozen paid services.