A benchmark comparing eight AI agent harnesses running the same Kimi K3 model across 25 business-application tasks. It compares pass rates, speed, tool calls, token usage, and estimated cost, finding substantial differences between harnesses.
Lightweight self-hosted app to track fuel, tax, insurance, servicing, and vehicle lifecycle costs. PWA for quick logging at the pump, with REST API and ready-to-use Home Assistant package (templates + reminders). Single FastAPI + SQLite app, Docker Compose ready with pre-built multi-arch images. 5 stars, MIT.
Practical guide to the Simple Memory node in n8n AI Agent. Covers context window (exchanges vs messages), sessionId configuration, queue mode incompatibility with Redis, and Postgres Chat Memory for production.
An autonomous AI agent that runs every weekday morning, scrapes LinkedIn for fresh job postings, scores them against your profile, and delivers matched jobs with ready-to-send cold emails directly to your Google Sheet and inbox.
Figranium is a self-hosted, free browser automation and scraping platform built on Playwright, designed to handle everything from simple scraping to complex, human-like browser interactions. It provides a visual task editor, a structured JSON task format, and advanced execution modes that go far beyond traditional scrapers.
Local-first browser automation platform with visual task builder, API access, and privacy-centric design.
Free tool to audit n8n workflows — detects exposed API keys, insecure webhooks, missing error handlers, expensive AI nodes, and infinite loops. 100% in-browser, no data leaves your device.
n8n queue mode build with automatic worker scaling and puppeteer
There is a massive amount of “he-said, she-said” in this sub about building a SaaS on top of n8n. Some people say you need to cough up for an Embed license the second you charge a customer. Others say you can do whatever you want if you self-host.