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

# Recommendations

> Read the records PolicyMate recommends to your workspaces, with their importance and the reason.

In addition to raw data, PolicyMate reads every new record against your workspaces and tells you which records matter and why. These judgements are called recommendations, and they are the same recommendations your team sees in the PolicyMate feed, digests, and alerts, so the API and the app always agree.

## Workspaces and Topics

A workspace holds a monitoring profile, which describes what your team cares about in plain language, along with the jurisdictions it covers. Within a workspace, topics name the themes you follow, such as "Data center siting". You may list the workspaces your API key can read, along with their topics, using the `GET /workspaces` and `GET /topics` endpoints.

## Retrieving Recommendations

The `GET /recommendations` endpoint returns the records recommended to your workspaces, newest first. It accepts the `workspace`, `topic`, `importance`, `record_type`, and `since` query parameters. For example, the following request retrieves bills recommended at `high` importance or above:

```http theme={null}
GET /api/v0/recommendations?importance=high&record_type=bill
```

A recommendation looks like the following:

```json theme={null}
{
  "id": "f02b...",
  "recommended_at": "2027-02-04T07:31:09Z",
  "workspace": { "id": "3a9d...", "name": "Data centers" },
  "topics": [ { "id": "6e1f...", "name": "Energy and water use" } ],
  "importance": "high",
  "why": "Creates new PUC reporting duties for data centers above 75 MW, which would cover most hyperscale sites in Texas.",
  "record": {
    "type": "bill",
    "id": "9b3e2c1a-...",
    "title": "HB 1234: Relating to the water and energy use of large data centers.",
    "url": "https://app.policymate.io/entity/9b3e2c1a-..."
  }
}
```

The `importance` attribute is one of `low`, `medium`, `high`, or `critical`. When filtering, the `importance` parameter acts as a minimum, so `high` also returns `critical` recommendations.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.