what it does
add your homelab gear once — servers, network, storage, the ups, the pi in the corner — plus the services running on each box. rackfolio rolls it into the numbers you can never remember off the top of your head:
- power draw while idle, and the estimated monthly cost at your $/kWh
- total invested across the rack
- rack units used of your total
- cores, ram, storage (raw and usable)
- what’s running across everything (proxmox, truenas, pfsense, plex…)
it’s private by default. later it becomes a profile you can share — the “what’s in my homelab” page you’d otherwise hand-write for the hundredth time.
power and cost are estimates until a gear catalog backs them (a later phase). the app says so, out loud, rather than pretending the watts are gospel.
why i built it
my own lab lives in a markdown file that’s out of date the week after i touch it. i forget which drives are in which bay, when the ups battery is due, what that third mini pc was even for. and i’ve never actually known my lab’s total power draw or monthly cost until something forced the math.
pcpartpicker exists for desktop builds. there’s no clean equivalent for a rack. rackfolio is that — manage it privately, see the totals automatically, and eventually show it with one link.
how i’m building it
one rule on rackfolio: ai helps me plan, name, and prototype, but it doesn’t write the application code. i’ve spent a long stretch building with ai in the loop, and i want to keep the part of me that actually builds things sharp. so the scoping, the naming, the design mockups — ai is fair game there. the code that ships is hand-written.
the plan
the ordering is the strategy: a genuinely good private tool first, sharing as a layer on top, community last. each phase is useful on its own.
- the tool — private inventory, the live stat block, devices and services. no auth, no sharing. i dogfood it on my own lab.
- profiles — accounts, claim a username, flip a lab public at
/u/<name>. the part you’d post to r/homelab. - the catalog — a seeded gear database so power and cost stop being hand-typed guesses and start being real.
- community — browse other people’s labs, filter by what they run, and the leaderboards i want to see: most powerful, best watts-per-core, biggest storage.
what’s next
right now the foundation is down — next.js scaffold, a dockerized postgres, and the screens designed as static mockups (that’s the hero above). next up is the react port and the data layer: the prisma schema, create/edit devices and services, and the live stat block running against a real database.