# Create semantic entries

Create semantic configuration entries to supplement data tables with definitions of fields, indicators, associations, and business rules.

```text
POST https://moi.matrixorigin.cn/newmoi/semantic-models/{model_id}/entries
```

## Preparation before calling

First [query the knowledge base list](list-knowledge-bases.md) to obtain the knowledge base ID. Prepare a personal access token and target workspace ID that has 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.
- `$MODEL_ID`: ID of the knowledge base for which the semantic entry is to be created.

## Request example

```bash
curl -X POST "https://moi.matrixorigin.cn/newmoi/semantic-models/$MODEL_ID/entries" \
  -H "X-API-Key: $AI_STUDIO_API_KEY" \
  -H "X-Workspace-ID: $WORKSPACE_ID" \
  -H 'Content-Type: application/json' \
  -d '{
    "kind": "metric",
    "key": "total_rows",
    "tables": ["orders"],
    "spec": {
      "expr": "COUNT(*)"
    }
  }'
```

## Path parameters

| Parameters | Type | Description |
| --- | --- | --- |
| `model_id` | integer | The knowledge base ID for which the semantic entry is to be created. |

## Request body

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `kind` | string | Yes | Entry type; it determines the structure of `spec`. |
| `key` | string | Yes | A stable reference key within the knowledge base. |
| `tables` | array[string] | No | The associated table name, not the Catalog table ID. |
| `spec` | object | Yes | A type-specific configuration object; fields are determined by `kind`. |

`key` must be unique in the same knowledge base; `tables` can be omitted when not passed. When passed in, each name must be a configured table of the knowledge base.

### `spec` configuration

In the same entry, first select `kind` and then submit the corresponding `spec` field according to the table below.

| Console Name | `kind` | `spec` Required Field | Purpose |
| --- | --- | --- | --- |
| Dimension column | `dimension` | `column` (string) | Describes the fields used for grouping and filtering. |
| Fact column | `fact` | `column` (string) | Describes the fact fields that can participate in the calculation. |
| Business indicators | `metric` | `expr` (string) | Define the indicator calculation expression. |
| Table association | `relationship` | `left_table` (string), `right_table` (string), `join_columns` (object[]) | Define the association of two configured tables. |
| Column preferences | `column_preference` | `preferred` (string), `deprecated` (string) | Specify recommended and non-recommended columns. |
| Named filter | `named_filter` | `expr` (string) | Define a reusable filter expression. |
| Default constraints | `default_constraint` | `column` (string), `values` (string[]) | Defines the default filter constraints for the column; only supports configuration through API. |
| Standard Q&A | `verified_query` | `question` (string), `sql` (string) | Save high-frequency questions and their SQL. |
| Terminology explanation | `glossary` | `term` (string), `definition` (string) | Define business terms and their meanings. |
| Rule injection | `logic_text` | `content` (string), `injection_stages` (string[]) | Apply business rules at the specified stage. |
| SQL result set | `sql_resultset` | `sql` (string), `description` (string) | Save SQL result set definition. |

`[]` after a type means an array. For example, `string[]` is an array of strings.

### Type-specific constraints

| `kind` | Field or condition | Constraint |
| --- | --- | --- |
| `relationship` | `join_columns` | The array cannot be empty; each item must contain a non-empty `left` and `right` (string). |
| `column_preference` | `preferred`, `deprecated` | Neither value can be empty, and cannot be the same without case sensitivity. |
| `default_constraint` | `values` | Contains at least one non-empty string; optional `operator` is `=`, `!=`, `<>`, `IN`, or `NOT IN`. |
| `logic_text` | `injection_stages` | The array cannot be empty; each item can only be `planner_policy`, `sql_generation`, `sql_followup`, `sql_regenerate`, `sql_decomposition`, `executor_rule` or `renderer_rule`. |
| `sql_resultset` | `sql`, `description` | `sql` 16 KiB or less; `description` 1000 characters or less. |

#### `sql_resultset` optional configuration

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `resolve_mode` | string | No | Can only be `semantic` or `passthrough`. `expand_sql` cannot be set when set to `passthrough`. |
| `max_rows` | integer | No | Cannot be less than `0`. |
| `max_bytes` | integer | No | Cannot be less than `0`. |
| `timeout_seconds` | integer | No | Range is `0` to `60`. |
| `expand_sql` | object | No | SQL expansion configuration. Both `expand_sql.sql` and `expand_sql.params` must be submitted during setup. |
| `expand_sql.sql` | string | Conditionally required | No more than 16 KiB; `{{parameter_name}}` must correspond to `expand_sql.params` one-to-one. |
| `expand_sql.params` | array[string] | Conditionally required | Quantity is `1` to `8`. Each name begins with a letter or underscore and can only contain letters, numbers, or underscores; it cannot be repeated if it is case-insensitive, and must be used in `expand_sql.sql`. |
| `retrieval` | object | No | Retrieve configuration. |
| `retrieval.enabled` | boolean | No | When set to `true`, `retrieval.embedding_model` must also be submitted. |
| `retrieval.embedding_model` | string | Conditionally required | `retrieval.enabled` must be a non-empty string when it is `true`. |

## Successful response

Return `201` on success; save `data.id` for subsequent updates and deletions.

```json
{
  "code": "OK",
  "msg": "OK",
  "data": {
    "id": 501,
    "kind": "metric",
    "key": "total_rows",
    "tables": ["orders"],
    "spec": {
      "expr": "COUNT(*)"
    },
    "created_at": 1735632000,
    "updated_at": 1735632000
  }
}
```

The response fields are as follows.

| Field | Type | Description |
| --- | --- | --- |
| `code` | string | `OK` on success. |
| `msg` | string | `OK` on success. |
| `data.id` | integer | New entry ID. |
| `data.kind` | string | The type of entry created. |
| `data.key` | string | Stable reference key within the knowledge base. |
| `data.tables` | array[string] | Association table name; may be omitted if not set. |
| `data.spec` | object | Saved type-specific configuration object; fields determined by `data.kind`. |
| `data.created_at` | integer | Unix timestamp. |
| `data.updated_at` | integer | Unix timestamp. |

## Error response

```json
{
  "code": "ErrParamInvalid",
  "msg": "invalid semantic entry",
  "data": null
}
```

### Common HTTP errors

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

* - HTTP status code
  - error code
  - Common causes
  - Recommended actions
* - `400`
  - `ErrParamInvalid`
  - `kind`, `key` or `spec` does not comply with the entry type contract.
  - Fix the complete entry definition and try again.
* - `401`
  - `ErrUnauthorized`
  - The API Key is invalid or has expired.
  - Check API Key.
* - `403`
  - `ErrForbidden`
  - The caller does not have permission to create entries.
  - Check workspace and object authorization.
* - `404`
  - `ErrNotFound`
  - The knowledge base does not exist or is not visible.
  - Reconfirm `model_id`.
* - `409`
  - `ErrConflict`
  - A conflicting entry key already exists in the same knowledge base.
  - Use a different `key` or update an existing entry.
* - `500`
  - `ErrServer`
  - The service failed to create the entry.
  - Keep the desensitized response information and try again.
```

## Follow-up operations

Use `data.id` [Update semantic entries](update-semantic-entry.md) or [Delete semantic entries](delete-semantic-entry.md).