Most writing about AI agents is either hype or homework. This is neither. It is a filterable directory of concrete things you can point an agent at, written for people who just got one and are wondering what to do with it.
Each use case points to a working reference (an n8n template, a Make blueprint, a real build), lists the tools it needs, and is dated when it was last checked. The trust label on the page says how much checking stands behind it.
Browse by role (what is your job?), by pattern (what shape of work is this?), or by tool (what do you already own?). The coverage matrix shows where the gaps are, honestly.
I am Mario Berger. I design, build, and run AI agents for small teams, and I keep this directory as the resource I wish existed when I started. If you want one of these set up for your business, I am one message away.
Every entry is curated from public implementations — workflow templates, vendor case studies, community write-ups — and from reader submissions. Each entry links its sources, so you can read the original and judge the fit yourself.
Three labels say how much checking stands behind an entry:
We do not test every setup ourselves. Most entries document someone else’s implementation, edited for clarity. Treat them as starting points with sources, not as guarantees.