02 — Own your infrastructure

Build your clone.
It works like you.

Your process is your edge — your checks, your releases, your judgment. Encode it into playbooks with triggers you define; let your agents build them, run them, and report back in tables and notes you own. Local AI infrastructure that compounds, on your machine.

  1. Apps: Apps that keep themselves up to date.

    Your agents build the app; its own schedule keeps it fresh. Double-click into one and it's all there: its pages, tables, playbooks and triggers, and an editor for every part of it.

  2. Calendar: Your meetings and your agents, on one calendar.

    Schedule a playbook run or a message to an Argus, once or on repeat. Connect the calendars already on your Mac and your week shows up beside them. Each App keeps its own calendar for its jobs.

  3. Playbooks: Deterministic steps. AI where you need judgement.

    Chain scripts, HTTP calls and agent steps into a run you can trust. Ask an agent a question mid-run, and its answer picks the next step. Your agent can write the playbook, and you can add to it, one cell at a time.

  4. Tables: Your own local database. No server, no signup.

    Open any row in a panel, drag it across a kanban, chart it in Visual, and change its columns as you go. One table, every view, all on your machine.

  5. Integrations: Your tools, grouped your way.

    Click a pill and a flat list of GitHub issues nests by repository, then by label. Add Jira and do it again. Your credentials stay on your Mac and are only ever sent to the service they belong to.

  6. Triggers: When your integration changes, the right thing runs.

    Point a trigger at the Jira you just added. When a new record matches your conditions, it runs a playbook or prompts an Argus, within the limits you set.

  7. Notes: Just start typing. @ links everything else.

    Headings, checklists and a slash menu, like the docs you already know. Type @ to mention a table, a playbook, an integration or another note, and one click opens it.