Digital Testing

Test chat bots the way you test IVRs: hold real conversations with a bot through a web widget, Amazon Connect chat or an HTTP endpoint, and get a verdict with the evidence behind it.

18 endpoints

BetaAvailable to beta workspaces only

Digital testing is in beta. Access is limited to beta workspaces, and these endpoints may change before general release. The SDKs cover runs, configs and compliance audits; connectors are reached through the API directly for now.

Run a digital test

POST/digital-testing/runs

Beta: available to beta workspaces only. Queue chat conversations with a bot and get a verdict. Returns 202 with a `pending` run; poll Get a digital test run. This costs money: each sample is a real conversation with your bot, billed in AI tokens, and the judged kinds run at least 3 samples. Three kinds: `standard` (did it say the right thing - step assertions, no model), `freeform` (did the customer get what they came for - judged against `acceptance`) and `compliance` (did the bot behave - judged against a catalogue test). Point it at a saved connector with `connectorId`, or send `target` and `targetRef` together. 402 when the workspace is not entitled to digital testing; 429 when 25 runs are already in flight.

Body Parameters

Name	Type	Required	Description
connectorId	string	optional	The connector to run against. Preferred over an inline target; resolved when the run starts, so editing the connector updates every test that uses it
targetRef	string	optional	Inline target identity reports group on, e.g. acme/billing-bot. Required with target when there is no connectorId
target	object	optional	Inline target, sent with targetRef when there is no connectorId. Discriminated on transport
└ transport	string	required	web-widget, connect-chat or http-json
└ url	string	optional	web-widget: the page the chat widget is embedded on
└ instanceId	string	optional	connect-chat: Amazon Connect instance ID
└ contactFlowId	string	optional	connect-chat: contact flow ID
└ endpoint	string	optional	http-json: the chat endpoint URL
sector	string	required	e.g. utilities
mission	string	required	What the customer is trying to do, e.g. Pay my bill
kind	string	optional	freeform (judged against acceptance), compliance (judged against a catalogue test) or standard (step assertions, no model). Default freeform
acceptance	string	optional	Required for freeform: what counts as success
catalogueTestId	string	optional	Required for compliance, e.g. M-001
passConditions	string[]	optional	All must hold for a pass. Seen by the judge only, never sent to the bot
failConditions	string[]	optional	Any one fails the test. Seen by the judge only
steps	ChatStep[]	optional	standard only: send, expect, wait and end steps. A judged kind with steps is a 400
probes	string[]	optional	Probe messages for the judged kinds
profileId	string	optional	A saved profile to draw test data from (see Profiles)
samples	number	optional	Conversations to run before the verdict, 1-10. Default 3 for the judged kinds (at least 3), and exactly 1 for standard
maxTurns	number	optional	Turn limit per conversation, 1-60. Default 12

Request

bash

curl -X POST -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"targetRef":"acme/support-widget","target":{"transport":"web-widget","url":"https://example.com/support"}, class="text-amber-300">"sector":"utilities","mission":"Check my outstanding balance","kind":"freeform","acceptance":"The bot states the outstanding balance."}' \
  https://api.nopaque.co.uk/digital-testing/runs

Response

json

{
  "message": "Digital test queued",
  "run": {
    "id": "6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b",
    "workspaceId": "ws_abc123",
    "targetRef": "acme/support-widget",
    "channel": "chat",
    "target": {
      "transport": "web-widget",
      "url": "https://example.com/support"
    },
    "kind": "freeform",
    "sector": "utilities",
    "mission": "Check my outstanding balance",
    "acceptance": "The bot states the outstanding balance.",
    "status": "pending",
    "samplesRequested": 3,
    "startedAt": "2026-11-04T14: 00: 00Z",
    "launchDeadline": "2026-11-04T14: 10: 00Z"
  }
}

List digital test runs

GET/digital-testing/runs

Beta: available to beta workspaces only. Runs in the workspace, newest first. Pass `targetRef` to list one target only. Page with `cursor` until `nextCursor` is absent.

Query Parameters

Name	Type	Required	Description
targetRef	string	optional	Only runs against this target
limit	number	optional	Page size, 1-200. Default 50. Outside that range is a 400
cursor	string	optional	nextCursor from the previous page

Request

bash

curl -H "x-api-key: YOUR_API_KEY" \
  "https://api.nopaque.co.uk/digital-testing/runs?targetRef=acme/support-widget&limit=50"

Response

json

{
  "runs": [
    {
      "id": "6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b",
      "workspaceId": "ws_abc123",
      "configId": "2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d",
      "targetRef": "acme/support-widget",
      "channel": "chat",
      "target": {
        "transport": "web-widget",
        "url": "https://example.com/support"
      },
      "kind": "freeform",
      "sector": "utilities",
      "mission": "Check my outstanding balance",
      "acceptance": "The bot states the outstanding balance.",
      "status": "completed",
      "outcome": "pass",
      "samplesRequested": 3,
      "samplesJudged": 3,
      "passed": 3,
      "failed": 0,
      "transportErrors": 0,
      "passRate": 1,
      "sampleOutcomes": [
        "pass",
        "pass",
        "pass"
      ],
      "samples": [
        {
          "runId": "6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b",
          "kind": "freeform",
          "outcome": "pass",
          "transcript": [
            {
              "role": "customer",
              "text": "Hi, how much do I owe at the moment?",
              "at": "2026-11-04T14: 00: 05Z"
            },
            {
              "role": "bot",
              "text": "Your outstanding balance is £42.10.",
              "at": "2026-11-04T14: 00: 07Z"
            }
          ],
          "reasoning": "The bot stated the outstanding balance in its first reply.",
          "passEvidence": [
            {
              "condition": "Stated the outstanding balance",
              "quote": "Your outstanding balance is £42.10."
            }
          ],
          "failEvidence": [],
          "botTurns": 1
        }
      ],
      "startedAt": "2026-11-04T14: 00: 00Z",
      "completedAt": "2026-11-04T14: 01: 30Z"
    }
  ],
  "nextCursor": "eyJrIjoiLi4uIn0"
}

Get a digital test run

GET/digital-testing/runs/{id}

Beta: available to beta workspaces only. One run, with each sample's transcript, verdict and evidence once it finishes. `status` ends as `completed` (a verdict was reached, pass or fail), `failed` (the test could not be delivered; `failureReason` says why) or `cancelled`; stop polling on any of the three. A bot that behaved badly is `completed` with `outcome: fail`. `passRate` is passed over `samplesJudged`; `transportErrors` counts conversations that never happened and is kept separate.

Path Parameters

Name	Type	Required	Description
id	string	required	Digital test run ID

Request

bash

curl -H "x-api-key: YOUR_API_KEY" \
  https://api.nopaque.co.uk/digital-testing/runs/6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b

Response

json

{
  "id": "6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b",
  "workspaceId": "ws_abc123",
  "configId": "2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d",
  "targetRef": "acme/support-widget",
  "channel": "chat",
  "target": {
    "transport": "web-widget",
    "url": "https://example.com/support"
  },
  "kind": "freeform",
  "sector": "utilities",
  "mission": "Check my outstanding balance",
  "acceptance": "The bot states the outstanding balance.",
  "status": "completed",
  "outcome": "pass",
  "samplesRequested": 3,
  "samplesJudged": 3,
  "passed": 3,
  "failed": 0,
  "transportErrors": 0,
  "passRate": 1,
  "sampleOutcomes": [
    "pass",
    "pass",
    "pass"
  ],
  "samples": [
    {
      "runId": "6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b",
      "kind": "freeform",
      "outcome": "pass",
      "transcript": [
        {
          "role": "customer",
          "text": "Hi, how much do I owe at the moment?",
          "at": "2026-11-04T14: 00: 05Z"
        },
        {
          "role": "bot",
          "text": "Your outstanding balance is £42.10.",
          "at": "2026-11-04T14: 00: 07Z"
        }
      ],
      "reasoning": "The bot stated the outstanding balance in its first reply.",
      "passEvidence": [
        {
          "condition": "Stated the outstanding balance",
          "quote": "Your outstanding balance is £42.10."
        }
      ],
      "failEvidence": [],
      "botTurns": 1
    }
  ],
  "startedAt": "2026-11-04T14: 00: 00Z",
  "completedAt": "2026-11-04T14: 01: 30Z"
}

Cancel a digital test run

POST/digital-testing/runs/{id}/cancel

Beta: available to beta workspaces only. Mark a run cancelled. A conversation already under way may still finish and write its evidence, but the run stays `cancelled`. 400 when the run has already finished.

Path Parameters

Name	Type	Required	Description
id	string	required	Digital test run ID

Request

bash

curl -X POST -H "x-api-key: YOUR_API_KEY" \
  https://api.nopaque.co.uk/digital-testing/runs/6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b/cancel

Response

json

{
  "id": "6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b",
  "workspaceId": "ws_abc123",
  "targetRef": "acme/support-widget",
  "channel": "chat",
  "target": {
    "transport": "web-widget",
    "url": "https://example.com/support"
  },
  "kind": "freeform",
  "sector": "utilities",
  "mission": "Check my outstanding balance",
  "acceptance": "The bot states the outstanding balance.",
  "status": "cancelled",
  "samplesRequested": 3,
  "startedAt": "2026-11-04T14: 00: 00Z",
  "launchDeadline": "2026-11-04T14: 10: 00Z"
}

Create a digital test config

POST/digital-testing/configs

Beta: available to beta workspaces only. Save a reusable test definition, then launch it with Launch a digital test config. The rules for each kind are checked when you save, so a config that could never run is a 400 rather than stored. Whether a `connectorId` exists is only checked at launch, before anything is billed.

Body Parameters

Name	Type	Required	Description
name	string	required	Up to 200 characters
description	string	optional	Up to 2000 characters
connectorId	string	optional	The connector to run against. Preferred over an inline target; resolved when the run starts, so editing the connector updates every test that uses it
targetRef	string	optional	Inline target identity reports group on, e.g. acme/billing-bot. Required with target when there is no connectorId
target	object	optional	Inline target, sent with targetRef when there is no connectorId. Discriminated on transport
└ transport	string	required	web-widget, connect-chat or http-json
└ url	string	optional	web-widget: the page the chat widget is embedded on
└ instanceId	string	optional	connect-chat: Amazon Connect instance ID
└ contactFlowId	string	optional	connect-chat: contact flow ID
└ endpoint	string	optional	http-json: the chat endpoint URL
sector	string	required	e.g. utilities
mission	string	required	What the customer is trying to do, e.g. Pay my bill
kind	string	optional	freeform (judged against acceptance), compliance (judged against a catalogue test) or standard (step assertions, no model). Default freeform
acceptance	string	optional	Required for freeform: what counts as success
catalogueTestId	string	optional	Required for compliance, e.g. M-001
passConditions	string[]	optional	All must hold for a pass. Seen by the judge only, never sent to the bot
failConditions	string[]	optional	Any one fails the test. Seen by the judge only
steps	ChatStep[]	optional	standard only: send, expect, wait and end steps. A judged kind with steps is a 400
probes	string[]	optional	Probe messages for the judged kinds
profileId	string	optional	A saved profile to draw test data from (see Profiles)
samples	number	optional	Conversations to run before the verdict, 1-10. Default 3 for the judged kinds (at least 3), and exactly 1 for standard
maxTurns	number	optional	Turn limit per conversation, 1-60. Default 12

Request

bash

curl -X POST -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"name":"Balance enquiry","targetRef":"acme/support-widget","target":{"transport":"web-widget","url":"https://example.com/support"}, class="text-amber-300">"sector":"utilities","mission":"Check my outstanding balance","kind":"freeform","acceptance":"The bot states the outstanding balance."}' \
  https://api.nopaque.co.uk/digital-testing/configs

Response

json

{
  "id": "2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d",
  "workspaceId": "ws_abc123",
  "name": "Balance enquiry",
  "targetRef": "acme/support-widget",
  "target": {
    "transport": "web-widget",
    "url": "https://example.com/support"
  },
  "sector": "utilities",
  "mission": "Check my outstanding balance",
  "kind": "freeform",
  "acceptance": "The bot states the outstanding balance.",
  "samples": 3,
  "createdAt": "2026-11-03T09: 00: 00Z",
  "updatedAt": "2026-11-03T09: 00: 00Z"
}

List digital test configs

GET/digital-testing/configs

Beta: available to beta workspaces only. Saved test definitions, most recently updated first. Page with `cursor` until `nextCursor` is absent.

Query Parameters

Name	Type	Required	Description
limit	number	optional	Page size, 1-200. Default 50. Outside that range is a 400
cursor	string	optional	nextCursor from the previous page

Request

bash

curl -H "x-api-key: YOUR_API_KEY" \
  "https://api.nopaque.co.uk/digital-testing/configs?limit=50"

Response

json

{
  "configs": [
    {
      "id": "2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d",
      "workspaceId": "ws_abc123",
      "name": "Balance enquiry",
      "targetRef": "acme/support-widget",
      "target": {
        "transport": "web-widget",
        "url": "https://example.com/support"
      },
      "sector": "utilities",
      "mission": "Check my outstanding balance",
      "kind": "freeform",
      "acceptance": "The bot states the outstanding balance.",
      "samples": 3,
      "createdAt": "2026-11-03T09: 00: 00Z",
      "updatedAt": "2026-11-03T09: 00: 00Z"
    }
  ]
}

Get a digital test config

GET/digital-testing/configs/{id}

Beta: available to beta workspaces only. One saved definition, including its steps or probes.

Path Parameters

Name	Type	Required	Description
id	string	required	Digital test config ID

Request

bash

curl -H "x-api-key: YOUR_API_KEY" \
  https://api.nopaque.co.uk/digital-testing/configs/2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d

Response

json

{
  "id": "2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d",
  "workspaceId": "ws_abc123",
  "name": "Balance enquiry",
  "targetRef": "acme/support-widget",
  "target": {
    "transport": "web-widget",
    "url": "https://example.com/support"
  },
  "sector": "utilities",
  "mission": "Check my outstanding balance",
  "kind": "freeform",
  "acceptance": "The bot states the outstanding balance.",
  "samples": 3,
  "createdAt": "2026-11-03T09: 00: 00Z",
  "updatedAt": "2026-11-03T09: 00: 00Z"
}

Update a digital test config

PATCH/digital-testing/configs/{id}

Beta: available to beta workspaces only. Partial update: send only the fields to change. The merged config is checked, not the change alone, so an edit that would leave a test that cannot run (switching `kind`, say) is a 400.

Path Parameters

Name	Type	Required	Description
id	string	required	Digital test config ID

Body Parameters

Name	Type	Required	Description
name	string	optional	Up to 200 characters
description	string	optional	Up to 2000 characters
connectorId	string	optional	The connector to run against. Preferred over an inline target; resolved when the run starts, so editing the connector updates every test that uses it
targetRef	string	optional	Inline target identity reports group on, e.g. acme/billing-bot. Required with target when there is no connectorId
target	object	optional	Inline target, sent with targetRef when there is no connectorId. Discriminated on transport
└ transport	string	required	web-widget, connect-chat or http-json
└ url	string	optional	web-widget: the page the chat widget is embedded on
└ instanceId	string	optional	connect-chat: Amazon Connect instance ID
└ contactFlowId	string	optional	connect-chat: contact flow ID
└ endpoint	string	optional	http-json: the chat endpoint URL
sector	string	optional	e.g. utilities
mission	string	optional	What the customer is trying to do, e.g. Pay my bill
kind	string	optional	freeform (judged against acceptance), compliance (judged against a catalogue test) or standard (step assertions, no model). Default freeform
acceptance	string	optional	Required for freeform: what counts as success
catalogueTestId	string	optional	Required for compliance, e.g. M-001
passConditions	string[]	optional	All must hold for a pass. Seen by the judge only, never sent to the bot
failConditions	string[]	optional	Any one fails the test. Seen by the judge only
steps	ChatStep[]	optional	standard only: send, expect, wait and end steps. A judged kind with steps is a 400
probes	string[]	optional	Probe messages for the judged kinds
profileId	string	optional	A saved profile to draw test data from (see Profiles)
samples	number	optional	Conversations to run before the verdict, 1-10. Default 3 for the judged kinds (at least 3), and exactly 1 for standard
maxTurns	number	optional	Turn limit per conversation, 1-60. Default 12

Request

bash

curl -X PATCH -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"samples":5}' \
  https://api.nopaque.co.uk/digital-testing/configs/2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d

Response

json

{
  "id": "2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d",
  "workspaceId": "ws_abc123",
  "name": "Balance enquiry",
  "targetRef": "acme/support-widget",
  "target": {
    "transport": "web-widget",
    "url": "https://example.com/support"
  },
  "sector": "utilities",
  "mission": "Check my outstanding balance",
  "kind": "freeform",
  "acceptance": "The bot states the outstanding balance.",
  "samples": 5,
  "createdAt": "2026-11-03T09: 00: 00Z",
  "updatedAt": "2026-11-04T10: 00: 00Z"
}

Delete a digital test config

DELETE/digital-testing/configs/{id}

Beta: available to beta workspaces only. Delete a saved definition. Runs already launched from it keep their `configId` and their evidence.

Path Parameters

Name	Type	Required	Description
id	string	required	Digital test config ID

Request

bash

curl -X DELETE -H "x-api-key: YOUR_API_KEY" \
  https://api.nopaque.co.uk/digital-testing/configs/2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d

Response

json

{
  "message": "Digital test config deleted"
}

Launch a digital test config

POST/digital-testing/configs/{id}/runs

Beta: available to beta workspaces only. Queue a run from a saved config and return 202, exactly as Run a digital test does, with the same billing, limits and response. This costs money: each sample is a real conversation, billed in AI tokens. Send an empty body to run it as saved, or override the target: `connectorId`, or `target` and `targetRef` together (overriding only one of the pair is a 400). The run carries the config's `configId`.

Path Parameters

Name	Type	Required	Description
id	string	required	Digital test config ID

Body Parameters

Name	Type	Required	Description
connectorId	string	optional	Run against a different connector than the config's
targetRef	string	optional	Override, together with target
target	object	optional	Inline target, sent with targetRef when there is no connectorId. Discriminated on transport
└ transport	string	required	web-widget, connect-chat or http-json
└ url	string	optional	web-widget: the page the chat widget is embedded on
└ instanceId	string	optional	connect-chat: Amazon Connect instance ID
└ contactFlowId	string	optional	connect-chat: contact flow ID
└ endpoint	string	optional	http-json: the chat endpoint URL
samples	number	optional	Override the sample count, 1-10

Request

bash

curl -X POST -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{}' \
  https://api.nopaque.co.uk/digital-testing/configs/2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d/runs

Response

json

{
  "message": "Digital test queued",
  "run": {
    "id": "6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b",
    "workspaceId": "ws_abc123",
    "targetRef": "acme/support-widget",
    "channel": "chat",
    "target": {
      "transport": "web-widget",
      "url": "https://example.com/support"
    },
    "kind": "freeform",
    "sector": "utilities",
    "mission": "Check my outstanding balance",
    "acceptance": "The bot states the outstanding balance.",
    "status": "pending",
    "samplesRequested": 3,
    "startedAt": "2026-11-04T14: 00: 00Z",
    "launchDeadline": "2026-11-04T14: 10: 00Z",
    "configId": "2a3b4c5d-6e7f-4a8b-9c0d-1e2f3a4b5c6d"
  }
}

List connector types

GET/digital-testing/connector-types

Beta: available to beta workspaces only. The kinds of bot a connector can reach, with the `config` fields each accepts. Read this before creating a connector. `crossAccount` appears only for a type that can run in your own AWS account, and says what your role must trust.

Request

bash

curl -H "x-api-key: YOUR_API_KEY" \
  https://api.nopaque.co.uk/digital-testing/connector-types

Response

json

{
  "types": [
    {
      "type": "web-widget",
      "label": "Web chat widget",
      "hint": "Browser-automated embedded chat widget - for targets with no API at all.",
      "formFields": [
        {
          "name": "url",
          "label": "Page URL",
          "type": "text",
          "required": true,
          "placeholder": "https://example.com/support",
          "hint": "The page the chat widget is embedded on, not the widget vendor."
        },
        {
          "name": "selectorProfile",
          "label": "Selector profile",
          "type": "text",
          "required": false
        }
      ]
    }
  ]
}

Create a connector

POST/digital-testing/connectors

Beta: available to beta workspaces only. Save how to reach a bot, once, and point tests at it with `connectorId`. `config` is the target without `transport`, checked against the fields List connector types gives for `type`. `type` and `targetRef` cannot be changed later; `targetRef` is what compliance reports group on.

Body Parameters

Name	Type	Required	Description
name	string	required	Up to 200 characters
description	string	optional	Up to 2000 characters
type	string	required	A type from List connector types, e.g. web-widget
targetRef	string	required	Stable identity reports group on, e.g. acme/support-widget. Up to 200 characters
config	object	required	The fields for this type, e.g. { "url": "https://example.com/support" } for web-widget

Request

bash

curl -X POST -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"name":"Support widget (production)","type":"web-widget","targetRef":"acme/support-widget","config":{"url":"https://example.com/support"}}' \
  https://api.nopaque.co.uk/digital-testing/connectors

Response

json

{
  "id": "8c7b6a5d-4e3f-4a2b-9c1d-0e9f8a7b6c5d",
  "workspaceId": "ws_abc123",
  "name": "Support widget (production)",
  "type": "web-widget",
  "targetRef": "acme/support-widget",
  "config": {
    "url": "https://example.com/support"
  },
  "typeUpdatedAt": "web-widget#2026-11-01T12: 00: 00Z",
  "createdAt": "2026-11-01T12: 00: 00Z",
  "updatedAt": "2026-11-01T12: 00: 00Z"
}

List connectors

GET/digital-testing/connectors

Beta: available to beta workspaces only. The workspace's saved connectors, most recently updated first. Page with `cursor` until `nextCursor` is absent.

Query Parameters

Name	Type	Required	Description
type	string	optional	Only connectors of this type. An unknown type is a 400
limit	number	optional	Page size, 1-200. Default 50. Outside that range is a 400
cursor	string	optional	nextCursor from the previous page

Request

bash

curl -H "x-api-key: YOUR_API_KEY" \
  "https://api.nopaque.co.uk/digital-testing/connectors?type=web-widget"

Response

json

{
  "connectors": [
    {
      "id": "8c7b6a5d-4e3f-4a2b-9c1d-0e9f8a7b6c5d",
      "workspaceId": "ws_abc123",
      "name": "Support widget (production)",
      "type": "web-widget",
      "targetRef": "acme/support-widget",
      "config": {
        "url": "https://example.com/support"
      },
      "typeUpdatedAt": "web-widget#2026-11-01T12: 00: 00Z",
      "createdAt": "2026-11-01T12: 00: 00Z",
      "updatedAt": "2026-11-01T12: 00: 00Z"
    }
  ]
}

Get a connector

GET/digital-testing/connectors/{id}

Beta: available to beta workspaces only. One saved connector. 404 when it is not in this workspace.

Path Parameters

Name	Type	Required	Description
id	string	required	Connector ID

Request

bash

curl -H "x-api-key: YOUR_API_KEY" \
  https://api.nopaque.co.uk/digital-testing/connectors/8c7b6a5d-4e3f-4a2b-9c1d-0e9f8a7b6c5d

Response

json

{
  "id": "8c7b6a5d-4e3f-4a2b-9c1d-0e9f8a7b6c5d",
  "workspaceId": "ws_abc123",
  "name": "Support widget (production)",
  "type": "web-widget",
  "targetRef": "acme/support-widget",
  "config": {
    "url": "https://example.com/support"
  },
  "typeUpdatedAt": "web-widget#2026-11-01T12: 00: 00Z",
  "createdAt": "2026-11-01T12: 00: 00Z",
  "updatedAt": "2026-11-01T12: 00: 00Z"
}

Update a connector

PATCH/digital-testing/connectors/{id}

Beta: available to beta workspaces only. Change a connector's `name`, `description` or `config`. Every test that uses it picks up the change on its next run. Trying to change `type` or `targetRef` is a 400. The merged `config` is checked, not the change alone.

Path Parameters

Name	Type	Required	Description
id	string	required	Connector ID

Body Parameters

Name	Type	Required	Description
name	string	optional	Up to 200 characters
description	string	optional	Up to 2000 characters
config	object	optional	Fields to change in the connection details

Request

bash

curl -X PATCH -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"config":{"url":"https://example.com/help"}}' \
  https://api.nopaque.co.uk/digital-testing/connectors/8c7b6a5d-4e3f-4a2b-9c1d-0e9f8a7b6c5d

Response

json

{
  "id": "8c7b6a5d-4e3f-4a2b-9c1d-0e9f8a7b6c5d",
  "workspaceId": "ws_abc123",
  "name": "Support widget (production)",
  "type": "web-widget",
  "targetRef": "acme/support-widget",
  "config": {
    "url": "https://example.com/help"
  },
  "typeUpdatedAt": "web-widget#2026-11-01T12: 00: 00Z",
  "createdAt": "2026-11-01T12: 00: 00Z",
  "updatedAt": "2026-11-05T08: 00: 00Z"
}

Delete a connector

DELETE/digital-testing/connectors/{id}

Beta: available to beta workspaces only. Delete a connector. Refused with 409 while a saved config still uses it; the message names those configs. Point them at another connector, or delete them, first.

Path Parameters

Name	Type	Required	Description
id	string	required	Connector ID

Request

bash

curl -X DELETE -H "x-api-key: YOUR_API_KEY" \
  https://api.nopaque.co.uk/digital-testing/connectors/8c7b6a5d-4e3f-4a2b-9c1d-0e9f8a7b6c5d

Response

json

{
  "message": "Connector deleted"
}

List digital compliance audits

GET/digital-testing/compliance-audits

Beta: available to beta workspaces only. One entry per target with at least one compliance run, newest first, in one response (not paginated).

Request

bash

curl -H "x-api-key: YOUR_API_KEY" \
  https://api.nopaque.co.uk/digital-testing/compliance-audits

Response

json

{
  "audits": [
    {
      "targetRef": "acme/billing-bot",
      "lastRunAt": "2026-11-04T15: 20: 00Z",
      "runCount": 7,
      "catalogueTestIds": [
        "M-001",
        "M-004"
      ]
    }
  ]
}

Get a digital compliance report

GET/digital-testing/compliance-audits/report

Beta: available to beta workspaces only. The compliance report for one chat target, in the same shape as the voice report (see Compliance), plus `channel` and `targetRef`. Each test's last run also carries `passRate` and `samplesJudged`, because a chat verdict is a majority over several conversations, and `transportErrors` for conversations that never happened. A run whose outcome was inconclusive gives no verdict rather than a fail. `targetRef` goes in the query string because it contains a slash.

Query Parameters

Name	Type	Required	Description
targetRef	string	required	The target to report on, e.g. acme/billing-bot
sector	string	optional	Only catalogue tests that apply to this sector

Request

bash

curl -H "x-api-key: YOUR_API_KEY" \
  "https://api.nopaque.co.uk/digital-testing/compliance-audits/report?targetRef=acme/billing-bot"

Response

json

{
  "workspaceId": "ws_abc123",
  "phoneNumber": "acme/billing-bot",
  "catalogueVersion": 2,
  "summary": {
    "ran": 1,
    "passed": 1,
    "failed": 0,
    "total": 1
  },
  "regulations": [
    {
      "key": "eu-ai-act",
      "name": "EU Artificial Intelligence Act",
      "jurisdiction": "eu",
      "url": "https://eur-lex.europa.eu/eli/reg/2024/1689/oj",
      "summary": {
        "ran": 1,
        "passed": 1,
        "failed": 0
      },
      "tests": [
        {
          "id": "M-001",
          "name": "AI Announcement Check",
          "summary": "Bot must self-identify as AI before substantive content.",
          "severity": "critical",
          "weight": "primary",
          "clauses": [
            "Article 50(1)",
            "Article 50(5)"
          ],
          "primaryFor": true,
          "lastRun": {
            "runId": "0d9e8f7a-6b5c-4d3e-8f2a-1b0c9d8e7f6a",
            "ranAt": "2026-11-04T15: 20: 00Z",
            "completedAt": "2026-11-04T15: 22: 10Z",
            "status": "completed",
            "verdict": "pass",
            "channel": "chat",
            "passRate": 1,
            "samplesJudged": 3,
            "samplesRequested": 3,
            "transportErrors": 0
          }
        }
      ]
    }
  ],
  "channel": "chat",
  "targetRef": "acme/billing-bot"
}