How it works
Once your agent is connected, give it access to the tool your work lives in — Linear, Trello, Jira, GitHub Issues, Notion — the same way you would for any other task, using that tool’s own connector or API token. Then ask for the import in plain language:Read my Linear issues for the current cycle, and create a card in alicein under the right enabler for each one. Use my labels to decide which enabler. Skip anything already closed.Your agent reads the source, decides where each item belongs in your plan, and creates the cards. You watch them appear on the map.
Give it your credentials for the other tool, never the other way round. Alice.In does not ask for, hold, or store tokens for your other services — the connection is between your agent and that tool, on your machine.
Worth saying in the prompt
- Where things should land. “Everything labelled
infragoes under the deployment enabler.” Otherwise it will guess, and guesses are what you will spend the evening correcting. - What to skip. Closed issues, someone else’s backlog, the graveyard column.
- How much to bring. The current cycle usually beats the entire history. You can always import more later.
- Whether to bring the discussion. Long comment threads can come across as a card’s description, or be left behind.
Other sources
The same approach works for anything your agent can read:- A document or spreadsheet — “turn this roadmap doc into phases and enablers.”
- A repository — issues, milestones, or a
TODO.md. - A conversation — paste the notes from a planning call and ask for cards.