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

# Memory

> Recall and analyze content from large contexts and files inside a single request

```json theme={null}
{
  "type": "registry",
  "name": "memory-expert"
}
```

<Note>
  Memory is a **core toolbox connector**, it is foundational to most agent workflows where the agent runtime needs to reason over large inputs without bloating the prompt.
</Note>

The **Memory** connector lets agents recall and analyze content from large contexts (long documents, attached files, transcripts, or other bulky inputs) that would not fit cleanly inside the agent runtime's working prompt. The agent runtime hands the bulky context to the Memory connector and asks targeted questions; the Memory connector searches and analyzes that content and returns the relevant facts.

<Warning>
  The Memory connector does **not** persist information across sessions, conversations, or context boundaries. It only operates over the context that has already been provided to the agent in the current request. If you need durable, cross-session memory, that has to be implemented at the application layer.
</Warning>

<Tip>
  Use the Memory connector when a request involves a large body of text or files that the agent runtime should not load into its own prompt (for example a long discharge summary, a multi-page guideline, or an attached transcript) and you only need a few targeted facts back.
</Tip>

## Capabilities

The Memory connector can:

* Search large bodies of context for information matching a query
* Analyze long documents or files and return targeted facts
* Surface relevant excerpts so the agent runtime can ground its synthesis
* Reduce the size of the agent runtime's working prompt by offloading bulky lookups

## Use cases

* Querying a long clinical document or attached file for specific facts
* Pulling targeted passages out of a large transcript or note
* Summarizing or extracting from inputs that are too large to keep in the agent runtime prompt
* Providing analyzed excerpts as evidence for a downstream connector

## How to invoke

Send the relevant query or high-level objective as the `text` parameter. Do not pass new raw facts to the Memory connector. Provide only the question or objective so it can search and analyze the existing in-request context itself.

For questions that can be answered directly from the agent runtime's own prompt without consulting bulky context, you can answer directly without invoking the Memory connector.