Action items become real tasks
Review the follow-ups that matter and put them on a board where they can be prioritized, moved, completed, or handed to an AI agent.
Move meeting action items out of the notes and onto a working task board. runnrr keeps the meeting source attached and lets Claude Code or Codex handle suitable follow-up work.
runnrr is for the step after meeting notes. With Granola, it can scan recent calls for action items and useful follow-ups, then put the ones you approve onto a personal Kanban board with the meeting context attached. You can also capture tasks from notes or selected text on your Mac.
The difference is what happens next. Instead of stopping at an extracted checklist, each task can carry the source note, linked files, and decisions that explain it. When the follow-up is suitable for AI, Claude Code or Codex can work from that context and you review the result.
Review the follow-ups that matter and put them on a board where they can be prioritized, moved, completed, or handed to an AI agent.
Keep the note and surrounding context with the card instead of copying a one-line action item into another app and losing why it exists.
For suitable work, Claude Code or Codex can read the card and context before drafting, researching, updating a document, or changing code.
Capture the meeting
Use Granola for the call if you want automatic scanning, or keep using your existing notes and capture a task manually from selected text.
Review the action items
runnrr surfaces Granola follow-ups for review. Keep the real commitments and useful mentions; dismiss the noise before it becomes board work.
Move the work forward
Prioritize the card yourself or let Claude Code or Codex work from the attached context when the task is a good fit for an agent.
No. runnrr is not an AI meeting recorder or note-taker. Granola can be the capture layer; runnrr is the follow-through layer that turns selected meeting work into source-attached tasks.
With Granola connected, runnrr can scan recent meeting notes and surface action items for review. You choose what becomes a real task before it lands on the working board.
You can. runnrr becomes useful when those items need to be prioritized, tracked across meetings, or handed to an AI agent with the original context still attached.
For tasks that fit an agent workflow, Claude Code or Codex can work from the card and attached context. You still choose the task and review the actions or output.
Keep tasks, notes, and source context together. Run the right card with Claude Code or Codex. Free for Mac. No sign-up.