Optio

Optio coordinates coding agents and automated work for teams using their own infrastructure. A shared feed tracks pull requests, recurring jobs, interactive sessions and agents that stay available between tasks.

• Pull request follow-through: Agents receive CI failures, merge conflicts and review feedback so they can continue a change; configured automatic merging squash-merges approved PRs when checks pass and closes the issue.
• Scheduled and event-driven work: Start jobs from schedules, webhooks, GitHub or Slack events, or tickets in GitHub, GitLab, Linear, Jira and Notion.
• Persistent agent teams: Give agents memory, inboxes and separate instructions so they can wake on messages or events and coordinate with each other.
• Runtime choice: Run Claude Code, OpenAI Codex, GitHub Copilot, Google Gemini, Cursor, OpenCode or OpenClaw, with a different agent vendor for review if needed.
• Session access and tools: Follow conversations and logs, control terminals through the browser or native iOS and Android apps, and assign MCP tools and service connections by repository and runtime.

Optio is open source under the MIT license. Its self-hosted platform uses Kubernetes, and sessions can execute in cluster pods or directories on paired machines using the machine's existing CLI login. Server secrets don't transfer to paired laptops. Model and provider choices depend on the selected agent runtime; hosting Optio yourself doesn't make those providers local.

Session sharing requires organization sign-in and grants both viewing and control, including access to credentials available in the session. Work within the same execution pool is mutually trusted; ordinary pods aren't a security boundary for hostile tenants.