$OPTIO is live now on @ponsdotfamily . Open-source orchestration for AI agents on your own infrastructure. Prompt in, pull request out.
CA: 0x26e6a61fd4d354122e2f571a1ea3a2844e1b9fd1
Web: https://t.co/oIgiGyfjcE
Different kinds of work. One way to run them.
A coding task, a scheduled report, a terminal on your laptop, an agent your team can message. In Optio, they all start in the same place.
Different kinds of work. One way to run them.
A coding task, a scheduled report, a terminal on your laptop, an agent your team can message. In Optio, they all start in the same place.
Your next workflow starts here.
git clone https://t.co/B6uwzpnz7l
cd optio
./scripts/setup-local.sh
Needs Docker Desktop with Kubernetes, Node 22+, pnpm 10+ and Helm.
Open source, MIT → https://t.co/oIgiGyfjcE
Run it where it makes sense.
Repeatable work → isolated pods in your cluster.
Hands-on work → Optio Local on your own laptop, using your existing CLI login.
Same triggers. Same feed.
Step away. Stay in the loop.
Native iOS and Android apps keep your workspace in reach: check the queue, reply to an agent, open a terminal, start new work.
Work that needs you shows up as a push alert or a Live Activity.
Keep a specialist on hand.
Persistent agents have a name, instructions and an inbox. They wake on messages, schedules or events, run one turn, and can message each other.
A dispatcher plus specialists is a swarm you control.
Make the routine run itself.
A weekday morning briefing. Nightly dependency updates. A PagerDuty webhook that starts incident triage.
Schedule or event → Run → Result, with logs, retries and cost for every run.
Your agent. Your choice.
Claude Code, OpenAI Codex, GitHub Copilot, Google Gemini, Cursor, OpenCode, OpenClaw, or a plain terminal.
Switch per session, per repo, or A/B two agents on the same ticket.
From issue to merged PR.
Label a ticket. Optio runs the agent in a git worktree, opens a PR and watches CI.
CI fails or a reviewer asks for changes? The agent resumes with the feedback.
Green and approved? Squash-merge.
Every piece of work in Optio answers five questions:
When: what starts it?
Where: pod or your machine?
Who: which agent?
What: the prompt
Then: exit, wait for you, or keep running
One form. One feed.
Most agent products want your repo in their cloud.
Optio runs in your Kubernetes cluster and on machines you pair. Your agents still call the model providers you choose, but the orchestration, logs and secrets stay with you.
Your cluster. Your rules.
Your agents. Your infrastructure. One workspace.
Optio is open-source orchestration for AI agent work. Turn a prompt into a pull request, put a workflow on a schedule, pick up a session from your phone.
Self-hosted. MIT licensed.
https://t.co/oIgiGyfjcE