5 projects

devon_ships

·1d ago

Cohort

I made self-hosted feature flags without the enterprise pricing

Self-hosted feature flags without the enterprise pricing or the sales call. Toggle features per user or by percentage rollout from a clean dashboard, with SDKs for a few languages and everything running in one Docker container so your data stays yours. The part I sweated was making percentage rollouts sticky: I hash the user id into the bucket so a given user does not flip-flop between variants on every request, which sounds obvious but is easy to get wrong and miserable to debug. It evaluates flags inline with no network hop using a local cache that refreshes in the background. Audit logs and scheduled rollouts are on the list.

backendsaasopen-source

33

theo_hacks

·6d ago

Threadbare

I made a minimalist forum that is just markdown and links

No upvotes, no algorithm, just chronological threads. Self-hosted as a single binary. I built it to host a tiny community without feeding it to an ad model.

open-sourcebackendRust

54

devon_ships

·1w ago

Beacon

I made a status page and incident timeline you actually own

A status page and incident timeline that you actually own. Incidents are just markdown files with frontmatter, so your whole incident history lives in git instead of a vendor's database, and you can diff it, review it in a PR, and move it anywhere. It serves a clean public page plus an RSS feed, runs as a single self-hosted binary, and has no per-seat pricing because it is open source. I kept it text-forward on purpose: a status page does not need a hero video or an animated globe, it needs to load instantly during an outage when your main site is already down. Scheduled maintenance windows and a tiny subscriber API are next.

open-sourcesaasbackend

22

devon_ships

·1w ago

Loopback

I built a tool to record, trim, and replay API mocks in one click

A proxy that records real responses and lets you replay them as fixtures. Great for demos and for tests that should not hammer a third party API. Exports to a flat folder you can commit.

backendapiDevTools

22

priya_codes

·2w ago

Quantize

I got small LLMs running nicely on a Raspberry Pi

A wrapper plus web UI around llama.cpp tuned for single board computers like the Raspberry Pi. It streams tokens over websockets and stays under the Pi's tiny memory budget using Q4 quantization and mmap, so a 3B model actually fits in 2GB. Tokens per second are slow, but for a home assistant that mostly answers short questions it is real-time enough to feel responsive. I wanted something that ran entirely on hardware I own: no API keys, no usage caps, and no sending my house's voice commands off to someone's cloud. Setup is a single script that auto-detects the board. Local Whisper for voice input is the next milestone.

AIopen-sourcebackend

31