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

# Workflows

> How AI agents use MemoryBase in practice.

## Plan in chat, build with an agent

The most common pattern. Think through your idea in a chat, then hand off the building to an agent.

<Steps>
  <Step title="Plan in Claude or ChatGPT">
    Describe what you want to build. Go back and forth until you have a clear plan. MemoryBase captures the conversation automatically.
  </Step>

  <Step title="Organize into a project">
    In the MemoryBase extension, add this conversation to a project. Add any related conversations or notes.
  </Step>

  <Step title="Hand off to your agent">
    ```
    Search my MemoryBase for the "Shopping Cart Feature" project
    and implement everything we planned.
    ```
  </Step>
</Steps>

## Research across tools, then build

<Steps>
  <Step title="Research with ChatGPT">
    Compare approaches or solutions. MemoryBase captures the analysis.
  </Step>

  <Step title="Decide in Claude">
    Settle on an approach. MemoryBase captures this too.
  </Step>

  <Step title="Build with an agent">
    Your agent searches MemoryBase for both conversations and builds accordingly.
  </Step>
</Steps>

## Day-to-day continuity

For longer projects, every conversation you have adds to your memory. When you pick up work the next day, your agent searches MemoryBase for yesterday's context instead of you re-explaining everything.