Skip to main content
Optio turns coding tasks into merged pull requests. Submit a task from the web UI, a GitHub Issue, or a Linear ticket, and Optio handles the rest: provisioning an isolated environment, running an AI agent, opening a PR, monitoring CI, triggering code review, auto-fixing failures, and merging when everything passes. The key difference from running an AI agent manually is the feedback loop. When CI fails, the agent is automatically resumed with the failure context. When a reviewer requests changes, the agent picks up the review comments and pushes a fix. When everything passes, the PR is squash-merged and the linked issue is closed. You describe the work; Optio drives it to completion.

​
Key concepts

Tasks are the unit of work in Optio. A task has a title, a description, a target repository, and an agent type. Tasks move through a defined lifecycle — from queued to provisioning to running to PR opened to completed — with automatic retries and resumption along the way. Agents are the AI models that do the coding work. Optio supports Claude Code (Anthropic) and OpenAI Codex. You configure which model and prompt template to use per repository, and Optio injects your task description into the agent’s context automatically. Pods are the isolated Kubernetes environments where agents run. Optio uses a pod-per-repo architecture: one long-lived pod per repository, with each task running in its own git worktree inside that pod. This means the repo is cloned once, tooling is installed once, and subsequent tasks start almost immediately. The feedback loop is what makes Optio different from a one-shot agent runner. After a PR is opened, Optio polls GitHub every 30 seconds to track CI status, review state, and merge readiness. CI failures, merge conflicts, and reviewer feedback all trigger automatic agent resumption with the relevant context attached.

​
What problems Optio solves

Running AI coding agents at scale involves a lot of manual orchestration that gets tedious fast:
  • Context switching — you submit a task to an agent, then have to monitor it, re-run it when CI fails, apply review feedback, and manually merge. Optio eliminates all of that.
  • Environment setup overhead — spinning up a fresh container per task is slow. Optio’s pod-per-repo model clones and sets up the repository once, then reuses that environment across tasks.
  • No visibility — most agent runners give you a log file after the fact. Optio streams live agent output to the dashboard, shows pipeline progress through each stage, and tracks cost per task.
  • Single-agent bottleneck — teams often need multiple tasks running concurrently across several repositories. Optio manages concurrency, queuing, and pod scaling automatically.
  • Ticket-to-PR friction — assigning a GitHub Issue or Linear ticket to Optio takes one click. The resulting PR is linked back to the ticket, and the ticket is closed on merge.

​
Explore

Quick Start

Get Optio running locally in minutes with Docker Desktop and Kubernetes.

How It Works

Follow a task from intake through to merged PR and understand each stage.

Managing Tasks

Create, prioritize, reorder, and bulk-manage coding tasks across your repositories.

Feedback Loop

See how Optio automatically handles CI failures, merge conflicts, and review feedback.