> ## Documentation Index
> Fetch the complete documentation index at: https://docs.ando.so/llms.txt
> Use this file to discover all available pages before exploring further.

# Turn a Jam into Linear work

> Convert spoken decisions into concise, reviewable Linear issues.

Use this Recipe after a Jam produces decisions that should become executable
work.

## Before you start

* Add an [agent built in Ando](/docs/build-agents-in-ando) to the conversation
  that contains the Jam.
* Open the agent in **Studio** -> **Agents**, then connect Linear under
  **Connected tools**. See [agent tool
  setup](/docs/build-agents-in-ando#connect-tools).
* Make sure the Jam has a transcript and the agent can access the conversation.

## Paste this prompt

```text theme={"system"}
Review the latest Jam in this conversation.

First summarize:
- decisions people explicitly made,
- open questions,
- follow-up work you inferred.

Then draft the smallest set of Linear issues needed to carry out the decisions.
For each issue, include source context, a suggested owner, and acceptance
criteria. Do not create anything until I approve the drafts.
```

## What to look for

The response should distinguish a decision someone actually made from a task
the agent inferred. Edit the drafts in Ando, then explicitly approve only the
issues you want created.

## Keep in mind

Ando can search Jam transcript segments only when you can access the underlying
conversation. A noisy or incomplete transcript may need a quick human
correction before the issues are useful.

Related: [Run a verified engineering loop](/recipes/verified-engineering-loop).