---
title: "Get Mind Working Set"
method: GET
path: "/api/mind/v2/minds/{mind_id}/working-set"
tags: ["mind-v2"]
---

# Get Mind Working Set

`GET /api/mind/v2/minds/{mind_id}/working-set`

## Request

- Base URL: `https://app.thinqos.com` (the document declares no server; this is the origin it was published from)
- URL: `https://app.thinqos.com/api/mind/v2/minds/{mind_id}/working-set`
- Auth: none declared

Fast, TTL-filtered working-set read for a Mind (design.md section 6).

Identity-scoped: the caller must own the Mind (self, or active agent
owner) via the same gate the perspective endpoints use. Kept lean and
on its own endpoint so the chat hot path can read "what is in mind
now" without the heavier perspective joins (section 12 latency note).
TTL-expired items are filtered out of the response. This is a pure
read: get_db does not commit the request session, so it must not
mutate. The physical drop of stale items happens on the next
attention write (WorkingSetService.update), not on this poll.

## Path parameters

- `mind_id` string, uuid, required

## Headers

- `x-impersonate-identity-id` string, nullable

## Response `200`

Successful Response

- V2WorkingSetResponse: TTL-pruned working set plus its last-write timestamp.
  - `mind_id` string, uuid, required
  - `updated_at` string, date-time, nullable, required
  - `items` V2WorkingSetEntry[], required
    - `ref_kind` 'entity' | 'proposition' | 'mind_holding', required
    - `ref_id` string, uuid, required
    - `score` number, required
    - `decayed_at` string, date-time, required
    - `pinned` boolean

## Other responses

- `422`: Validation Error

---

[API](https://skmtc.dev/thinqos/apis/thinqos-api.md) · [All operations](https://skmtc.dev/thinqos/apis/thinqos-api/llms.txt) · [OpenAPI document](https://skmtc.dev/thinqos/apis/thinqos-api/revisions/5fdd1072e877?raw)
