# Help center

Source: https://ceyo.ai/docs/signal/help-center

### Help center

Read the Ceyo help center, including the Business help, the Partner playbook and the Developer docs, to host it in your own help center or support tools. The content is the same for every workspace.

> **Access**
>
> Requires a workspace-scoped key with `workspace:read`. Project and location keys receive `403 forbidden`.

> **Keeping a copy in sync**
>
> `updated_at` is when an article was edited, which can be a few days before the edit is published. The simplest sync is a daily fetch of all articles, comparing each `updated_at` with the one you stored. If you use `updated_since`, go back a few days from your last sync. Articles that are no longer returned were removed.

> **Caching**
>
> Responses include an `ETag` header. Send it back in `If-None-Match`: when nothing changed you get `304 Not Modified` with no body, which is the cheapest daily check. On a `200`, store the new ETag and compare each article's `updated_at`. Send `Accept-Encoding: gzip` to receive compressed responses.

### Get help center

`GET /help`

Returns the help categories and the glossary.

#### Response envelope

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `categories` | HelpCategory\[\] |  | All categories. |
| `glossary` | GlossaryTerm\[\] |  | All glossary terms. |

#### HelpCategory

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `id` | string |  | Stable category identifier. |
| `product` | merchant \| playbook \| developer |  | Help section the category belongs to. See product on HelpArticle. |
| `name` | string |  | Display name. |
| `description` | string |  | One-line summary. |

#### GlossaryTerm

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `term` | string |  | The term. Unique, and used as its identifier. |
| `aliases` | string\[\] |  | Other spellings and plural forms of the term. |
| `article` | string \| null |  | ID of the article that explains the term in depth. |
| `definition` | string |  | One-sentence definition. |
| `created_at` | datetime |  | When the term was added. |
| `updated_at` | datetime |  | When the term was last edited. |

#### Request and response

```curl
curl --request GET \
  --url 'https://api.signal.ceyo.ai/v1/help' \
  --header 'Authorization: Bearer ceyo_platform_...'
```

```json
{
  "categories": [
    {
      "id": "getting-started",
      "product": "merchant",
      "name": "Getting started",
      "description": "What it is, how it works and the first week."
    }
  ],
  "glossary": [
    {
      "term": "Prompt",
      "aliases": [
        "prompts",
        "tracked prompt",
        "tracked prompts"
      ],
      "article": "prompt-vs-topic",
      "definition": "A prompt is one request a customer might type into an AI model, tracked for a business and asked exactly as written.",
      "created_at": "2026-10-26T09:35:34Z",
      "updated_at": "2026-10-26T09:35:34Z"
    }
  ]
}
```

#### Errors

#### Error response envelope

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `error` | Error |  | Structured error payload. |

#### Error

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `code` | string |  | Stable snake\_case code suitable for programmatic handling. |
| `message` | string |  | Human-readable explanation of the failure. |
| `details` | object \| array \| null |  | Structured validation or request context when available. |
| `request_id` | string |  | Identifier to provide when requesting support. |

```json
{
  "error": {
    "code": "validation_failed",
    "message": "One or more fields are invalid.",
    "details": [
      {
        "field": "name",
        "message": "must be present"
      }
    ],
    "request_id": "req_01K1GP6J8QQFZ4D2B6C5A9V3TS"
  }
}
```

#### Status codes

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `401` | invalid\_api\_key |  | The API key is absent, invalid, expired, or revoked. |
| `403` | forbidden \| workspace\_suspended \| workspace\_archived |  | The key is not workspace-scoped, lacks workspace:read, or the workspace is suspended or archived. |

### List help articles

`GET /help/articles`

Returns help articles in help center order, with optional filters.

#### Query parameters

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `product` | merchant \| playbook \| developer | Optional | Only return articles from one help section. |
| `category` | string | Optional | Only return articles from one category. Categories can change, so an unknown category returns no articles instead of an error. |
| `type` | question \| guide | Optional | Only return questions or only guides. |
| `updated_since` | datetime | Optional | Only return articles updated at or after this ISO 8601 time, such as 2026-11-01T00:00:00Z. URL-encode a + in a time zone offset. |
| `page` | integer | Optional; Default: 1 | One-based page number. |
| `per_page` | integer | Optional; Default: 100 | Articles per page, from 1 to 200. |

#### Response envelope

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `articles` | HelpArticle\[\] |  | Articles on this page. |
| `pagination` | Pagination |  | page, per\_page, total and total\_pages. |

#### HelpArticle

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `id` | string |  | Stable article identifier. |
| `category` | string |  | ID of the category, from Get help center. |
| `product` | merchant \| playbook \| developer |  | Help section: merchant is the Business help for businesses and their staff, playbook is the Partner playbook for selling and supporting the product, and developer is the Developer docs for integration teams. |
| `type` | question \| guide |  | question is a frequently asked question; guide is a longer walkthrough. Filter on type=question for an FAQ. |
| `audience` | both \| partners \| agencies |  | Who the article is written for. |
| `order` | integer |  | Position of the article within its category. |
| `title` | string |  | Article title. |
| `short` | string |  | The answer in one paragraph of plain text. |
| `detail` | string |  | The full article in Markdown, with ### subheadings, lists and bold text. References to other articles read "(See: Article title)". |
| `variants` | string\[\] |  | Other ways people ask the same question. Useful as keywords in your own search. |
| `related` | string\[\] |  | IDs of related articles. |
| `features` | string\[\] \| null |  | Product features the article is about, such as actions. Null when it applies to the whole product. |
| `playbook_note` | string \| null |  | Extra advice for partners who sell or support the product. |
| `developer_note` | string \| null |  | Extra technical detail for integration teams. |
| `created_at` | datetime |  | When the article was added. |
| `updated_at` | datetime |  | When the article was last edited. |

#### Request and response

```curl
curl --request GET \
  --url 'https://api.signal.ceyo.ai/v1/help/articles?product=merchant&type=question' \
  --header 'Authorization: Bearer ceyo_platform_...'
```

```json
{
  "articles": [
    {
      "id": "cadence",
      "category": "account",
      "product": "merchant",
      "type": "question",
      "audience": "both",
      "order": 2,
      "title": "How often are my prompts measured?",
      "short": "How often your prompts are measured depends on your package; ask our support team which cadence you have. A steady reading comes from enough different prompts per topic, more than from how often they run.",
      "detail": "Cadence is how often your prompts are run on the AI models. Each round is called a scan.\n\n### Why the number of prompts matters most\n\nAnswers vary from one run to the next, so a single prompt says little. ...",
      "variants": [
        "How often are results updated, and is that often enough?",
        "Is a weekly scan enough?"
      ],
      "related": [
        "reliability",
        "history",
        "why-many"
      ],
      "features": null,
      "playbook_note": "A package's cost grows with the number of answers it collects, so more prompts at a lower cadence is often better value than fewer prompts at a higher one.",
      "developer_note": null,
      "created_at": "2026-10-26T09:35:34Z",
      "updated_at": "2026-10-26T16:52:57Z"
    }
  ],
  "pagination": {
    "page": 1,
    "per_page": 100,
    "total": 1,
    "total_pages": 1
  }
}
```

#### Errors

#### Error response envelope

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `error` | Error |  | Structured error payload. |

#### Error

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `code` | string |  | Stable snake\_case code suitable for programmatic handling. |
| `message` | string |  | Human-readable explanation of the failure. |
| `details` | object \| array \| null |  | Structured validation or request context when available. |
| `request_id` | string |  | Identifier to provide when requesting support. |

```json
{
  "error": {
    "code": "validation_failed",
    "message": "One or more fields are invalid.",
    "details": [
      {
        "field": "name",
        "message": "must be present"
      }
    ],
    "request_id": "req_01K1GP6J8QQFZ4D2B6C5A9V3TS"
  }
}
```

#### Status codes

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `401` | invalid\_api\_key |  | The API key is absent, invalid, expired, or revoked. |
| `403` | forbidden \| workspace\_suspended \| workspace\_archived |  | The key is not workspace-scoped, lacks workspace:read, or the workspace is suspended or archived. |
| `422` | validation\_failed |  | A query parameter is missing or invalid. |

### Get help article

`GET /help/articles/{id}`

Returns one help article.

#### Path parameters

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `id` | string | Required | Article identifier, such as cadence. |

#### Response envelope

The article.

#### Request and response

```curl
curl --request GET \
  --url 'https://api.signal.ceyo.ai/v1/help/articles/cadence' \
  --header 'Authorization: Bearer ceyo_platform_...'
```

```json
{
  "article": {
    "id": "cadence",
    "category": "account",
    "product": "merchant",
    "type": "question",
    "audience": "both",
    "order": 2,
    "title": "How often are my prompts measured?",
    "short": "How often your prompts are measured depends on your package; ask our support team which cadence you have. A steady reading comes from enough different prompts per topic, more than from how often they run.",
    "detail": "Cadence is how often your prompts are run on the AI models. Each round is called a scan.\n\n### Why the number of prompts matters most\n\nAnswers vary from one run to the next, so a single prompt says little. ...",
    "variants": [
      "How often are results updated, and is that often enough?",
      "Is a weekly scan enough?"
    ],
    "related": [
      "reliability",
      "history",
      "why-many"
    ],
    "features": null,
    "playbook_note": "A package's cost grows with the number of answers it collects, so more prompts at a lower cadence is often better value than fewer prompts at a higher one.",
    "developer_note": null,
    "created_at": "2026-10-26T09:35:34Z",
    "updated_at": "2026-10-26T16:52:57Z"
  }
}
```

#### Errors

#### Error response envelope

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `error` | Error |  | Structured error payload. |

#### Error

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `code` | string |  | Stable snake\_case code suitable for programmatic handling. |
| `message` | string |  | Human-readable explanation of the failure. |
| `details` | object \| array \| null |  | Structured validation or request context when available. |
| `request_id` | string |  | Identifier to provide when requesting support. |

```json
{
  "error": {
    "code": "validation_failed",
    "message": "One or more fields are invalid.",
    "details": [
      {
        "field": "name",
        "message": "must be present"
      }
    ],
    "request_id": "req_01K1GP6J8QQFZ4D2B6C5A9V3TS"
  }
}
```

#### Status codes

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `401` | invalid\_api\_key |  | The API key is absent, invalid, expired, or revoked. |
| `403` | forbidden \| workspace\_suspended \| workspace\_archived |  | The key is not workspace-scoped, lacks workspace:read, or the workspace is suspended or archived. |
| `404` | not\_found |  | No article has this ID. |

### Search help articles

`GET /help/search`

Returns up to 10 articles that match a question or keywords, best match first.

#### Query parameters

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `q` | string | Required | What the person is looking for, up to 500 characters. |
| `product` | merchant \| playbook \| developer | Optional | Only return articles from one help section. |
| `category` | string | Optional | Only return articles from one category. Categories can change, so an unknown category returns no articles instead of an error. |
| `type` | question \| guide | Optional | Only return questions or only guides. |

#### Response envelope

Matching articles, best match first. Empty when nothing matches.

> **Matching**
>
> Matches words in the title, the other phrasings, the short answer and the full article, including common synonyms such as "rivals" for competitors, and forgives a one-letter typo such as "competitorss". Titles count most.

#### Request and response

```curl
curl --request GET \
  --url 'https://api.signal.ceyo.ai/v1/help/search?q=how%20often%20are%20prompts%20measured' \
  --header 'Authorization: Bearer ceyo_platform_...'
```

```json
{
  "articles": [
    {
      "id": "cadence",
      "category": "account",
      "product": "merchant",
      "type": "question",
      "audience": "both",
      "order": 2,
      "title": "How often are my prompts measured?",
      "short": "How often your prompts are measured depends on your package; ask our support team which cadence you have. A steady reading comes from enough different prompts per topic, more than from how often they run.",
      "detail": "Cadence is how often your prompts are run on the AI models. Each round is called a scan.\n\n### Why the number of prompts matters most\n\nAnswers vary from one run to the next, so a single prompt says little. ...",
      "variants": [
        "How often are results updated, and is that often enough?",
        "Is a weekly scan enough?"
      ],
      "related": [
        "reliability",
        "history",
        "why-many"
      ],
      "features": null,
      "playbook_note": "A package's cost grows with the number of answers it collects, so more prompts at a lower cadence is often better value than fewer prompts at a higher one.",
      "developer_note": null,
      "created_at": "2026-10-26T09:35:34Z",
      "updated_at": "2026-10-26T16:52:57Z"
    }
  ]
}
```

#### Errors

#### Error response envelope

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `error` | Error |  | Structured error payload. |

#### Error

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `code` | string |  | Stable snake\_case code suitable for programmatic handling. |
| `message` | string |  | Human-readable explanation of the failure. |
| `details` | object \| array \| null |  | Structured validation or request context when available. |
| `request_id` | string |  | Identifier to provide when requesting support. |

```json
{
  "error": {
    "code": "validation_failed",
    "message": "One or more fields are invalid.",
    "details": [
      {
        "field": "name",
        "message": "must be present"
      }
    ],
    "request_id": "req_01K1GP6J8QQFZ4D2B6C5A9V3TS"
  }
}
```

#### Status codes

| Name | Type | Details | Description |
| --- | --- | --- | --- |
| `401` | invalid\_api\_key |  | The API key is absent, invalid, expired, or revoked. |
| `403` | forbidden \| workspace\_suspended \| workspace\_archived |  | The key is not workspace-scoped, lacks workspace:read, or the workspace is suspended or archived. |
| `422` | validation\_failed |  | A query parameter is missing or invalid. |
