---
title: "Steer Agent Work Item"
method: POST
path: "/agent/work-items/{work_item_id}/steer"
tags: ["Agents"]
---

# Steer Agent Work Item

`POST /agent/work-items/{work_item_id}/steer`

## Request

- Base URL: `https://be.graph8.com/api/v1`
- URL: `https://be.graph8.com/api/v1/agent/work-items/{work_item_id}/steer`
- Auth: HTTP bearer

Redirect a RUNNING work item at its next safe step boundary.

Not mid-step - that is what makes steering safe at all. The instruction is
picked up when the agent reaches a boundary, so a steer sent during a long
tool call lands afterwards rather than corrupting it.

SAFE TO RETRY: `client_turn_id` is a caller-generated idempotency key, and
the same id twice comes back `status: "duplicate"` rather than queueing a
second instruction into a running agent. Check that field before assuming
a retry did something.

## Path parameters

- `work_item_id` string, required: Work item id (UUID).

## Headers

- `Authorization` string, nullable
- `X-Target-Org-Id` string, nullable

## Request body

- AgentWorkItemSteerRequest: Redirect a RUNNING work item at its next safe step boundary. `extra="forbid"` mirrors the app: an unknown key here is a typo in an instruction payload, and dropping it silently would let a caller believe they had steered when they had not.
  - `client_turn_id` string, uuid, required: Caller-generated idempotency key. REQUIRED, and the reason steering is safe to retry: the same id twice comes back `duplicate` rather than queueing a second instruction into a running agent.
  - `instruction` string, required: What to do differently. Applied at the next safe step boundary, not mid-step.

## Response `200`

Successful Response

- ApiResponseDict
  - `data` object, required
  - `pagination` PaginationMeta
    - `page` integer, required: Current page number (1-indexed)
    - `limit` integer, required: Items per page
    - `total` integer, required: Total number of items
    - `has_next` boolean, required: Whether there are more pages
    - `next_cursor` string, nullable: Opaque cursor for the next page; null when there are no more items. Pass it back as the `cursor` query param to fetch the following page without computing page numbers.
    - `total_available` integer, nullable: Search endpoints only: how many records really match, when known. Can be larger than `total`, which is capped at the 10,000 results that can be paged. For company search this is a distinct-company estimate. Null when unknown; absent outside search.
    - `total_is_lower_bound` boolean, nullable: Search endpoints only: true when more records match than `total` shows (the result set is larger than the 10,000 results that can be paged). Absent outside search.
    - `notes` string[], nullable: Search endpoints only: plain-language notes about the result, e.g. that more records match than can be paged, or that a reversed `between` range was put in order.

## Other responses

- `422`: Validation Error

---

[API](https://skmtc.dev/graph8/apis/graph8-developer-api.md) · [All operations](https://skmtc.dev/graph8/apis/graph8-developer-api/llms.txt) · [OpenAPI document](https://skmtc.dev/graph8/apis/graph8-developer-api/revisions/2e7cfa0ca921?raw)
