01 Runs on your hardware
The container is the product: Docker on a laptop, LXC on Proxmox, one config file. Providers are plugins — Anthropic, xAI, Google, or fully local via Ollama, vLLM, and llama.cpp. Your keys, your GPUs, your rules.
02 Remembers everything
Postgres-backed memory with hybrid full-text + vector search, summary compaction, and cross-agent recall. Ask what you decided three weeks ago; get the answer in milliseconds, not a scroll through transcripts.
rivet> memory_search "mesh CA decision?"
● 3 hits · hybrid fts+vector · 48ms
1. 06-12 one root CA, per-node certs
2. 06-13 rotation via update --mesh
3. 06-27 roster fix deployed fleet-wide03 A fleet, not a bot
Run one agent or several — different models, one shared memory. Agents delegate to each other over an mTLS mesh: the reasoning model plans, the fast one executes, the local one stays free. Same runtime on every node.
Five minutes to a running agent
$ npx rivetos init The wizard walks through providers, channels, and deployment. Getting started →