---
title: "Retrieval and context"
manual: "AI Assistant"
version: "main"
source: "UserGuide/RetrievalAndContext.rst"
rendered: "2026-10-03T14:54:45+00:00"
---

# Retrieval and context {#retrieval-and-context}

Retrieval is the step where the assistant searches indexed content. The quality
of answers depends strongly on retrieval configuration and metadata.

## Key retrieval settings {#key-retrieval-settings}

-   **max_retrieval_results**

    Maximum number of results contributed by this retriever.

-   **score_threshold**

    Minimum similarity score. A threshold of `0` usually means no explicit
    threshold.

-   **max_context_chunks**

    Maximum number of chunks passed into the prompt context. A value of `0`
    can be used by the implementation to mean no explicit chunk limit.

-   **max_context_characters**

    Maximum number of characters passed into the prompt context. A value of
    `0` can be used by the implementation to mean no explicit character limit.

-   **prompt_metadata_fields**

    Comma-separated list of metadata fields included in the prompt context and
    exposed as answer sources in the frontend response.

## Recommended metadata fields {#recommended-metadata-fields}

For most assistants, start with:

```text
title,url,source_type,source_identifier
```

For debugging and source attribution you can add project-specific fields such as
`page_uid`, `record_identifier`, `collection` or `chunk_index` if they are
available in the payload.