# Query session messages

Read saved messages in visible sessions. This interface is read-only; new messages are written by the A2A runtime called by the agent and are not created through this interface.

```text
GET https://moi.matrixorigin.cn/newmoi/workspaces/{workspace_id}/conversations/{conversation_id}/messages
```

## Preparation before calling

First [query session details](get-session.md) or [create session](create-session.md)] to obtain the session ID. Prepare the personal access token, target workspace ID, and session ID that have access to the target workspace.

The example below uses:

- `$AI_STUDIO_API_KEY`: The actual personal access token, passed through the `X-API-Key` Header.
- `$WORKSPACE_ID`: Target workspace ID, passed through `X-Workspace-ID` Header.
- `$CONVERSATION_ID`: The session ID of the message to be queried.
- `$AGENT_ID`: Optional agent filter ID.

## Request example

```bash
curl --get "https://moi.matrixorigin.cn/newmoi/workspaces/$WORKSPACE_ID/conversations/$CONVERSATION_ID/messages" \
  -H "X-API-Key: $AI_STUDIO_API_KEY" \
  -H "X-Workspace-ID: $WORKSPACE_ID" \
  --data-urlencode "agent_id=$AGENT_ID" \
  --data-urlencode "role=assistant" \
  --data-urlencode "limit=50"
```

## Path parameters

| Parameters | Type | Is it required | Description |
| --- | --- | --- | --- |
| `workspace_id` | string | Yes | Current workspace ID, must be consistent with `X-Workspace-ID`. |
| `conversation_id` | string | Yes | The session ID of the message to be queried. |

## Query parameters

| Parameters | Type | Is it required | Description |
| --- | --- | --- | --- |
| `agent_id` | string | No | Only queries sessions belonging to the specified agent. Does not filter by agent ID when not provided. |
| `agent_workspace_id` | string | No | ID of the workspace to which the agent belongs. When `agent_id` is specified without this parameter, the current workspace is used; only the current workspace or the system workspace can be specified. |
| `role` | string | No | Filter by message role. |
| `limit` | integer | No | Single return quantity. `50` when not provided or less than or equal to `0`, up to `200`. |
| `offset` | integer | No | Returns the offset. `0` when not provided or less than `0`. |

## Successful response

Returns `200` on success. `data.items` is a saved message on this page. It provides related information and content segmentation according to the session message sequence number.

```json
{
  "code": 0,
  "data": {
    "items": [
      {
        "id": "msg_01",
        "workspace_id": "ws_01",
        "conversation_id": "conv_01",
        "task_id": "task_01",
        "agent_id": "agent_01",
        "role": "assistant",
        "parts": [
          {
            "kind": "text",
            "text": "分析已完成"
          }
        ],
        "seq": 2,
        "created_at": "2026-01-02T15:04:05Z"
      }
    ],
    "total": 1,
    "limit": 50,
    "offset": 0
  }
}
```

The response fields are as follows.

| Field | Type | Description |
| --- | --- | --- |
| `code` | integer | `0` on success. |
| `data.items` | object[] | List of messages on this page. |
| `data.items[].id` | string | Message ID. |
| `data.items[].workspace_id` | string | ID of the workspace to which the message belongs. |
| `data.items[].conversation_id` | string | The session ID to which the message belongs. |
| `data.items[].task_id` | string | Associated task ID; does not return if there is no associated task. |
| `data.items[].agent_id` | string | Associated agent ID; not returned if there is no associated agent. |
| `data.items[].manifest_id` | string | Associated runtime manifest ID; not returned if there is no associated manifest. |
| `data.items[].role` | string | Message role. |
| `data.items[].parts` | object[] | Message content segments. The fields of each element are determined by the message segment type. |
| `data.items[].parent_message_id` | string | Parent message ID; does not return if there is no parent message. |
| `data.items[].seq` | integer | Message sequence number within the session. |
| `data.items[].created_at` | string | Message creation time, in RFC 3339 format. |
| `data.total` | integer | The total number of messages that meet the filter conditions. |
| `data.limit` | integer | The upper limit of the actual return quantity used this time. |
| `data.offset` | integer | The offset actually used this time. |

`[]` after a type denotes an array. `[]` in a field path denotes each item in an array.

## Error response

```json
{
  "code": 3,
  "message": "会话不存在"
}
```

### Common HTTP errors

```{list-table}
:header-rows: 1
:widths: 12 22 32 34

* - HTTP status code
  - error code
  - Common causes
  - Recommended actions
* - `400`
  - `2`（`INVALID_ARGUMENT`）
  - Invalid session ID, agent filter parameter, or pagination parameter.
  - Check path and query parameters.
* - `404`
  - `3`（`NOT_FOUND`）
  - The session does not exist, is not visible, or does not belong to the specified agent scope.
  - Confirm session ID and agent filters.
* - `500`
  - `1`（`INTERNAL`）
  - The server cannot query messages or generate message display data.
  - Record the request time and error message and try again.
* - `503`
  - `15`（`UNAVAILABLE`）
  - Session service is temporarily unavailable.
  - Try again later.
```

## Follow-up operations

Use [Query session details](get-session.md) to read the session's current display information and active task associations.