@graph8/sdk 0.251.0 → 0.262.0

package/docs/api/documents.md

@@ -0,0 +1,1440 @@
+<!-- Generated by sdk/js/scripts/generate-agent-docs.ts from developer-contract/openapi.json. Regenerated on every build. -->
+
+# `g8.api.documents`: Documents
+
+35 operations (10 read, 20 write, 4 destructive, 1 billable).
+
+No hand-written client covers this area. Call it through `g8.api.documents.*` as shown below, or by operationId with `g8.api.call(operationId, input)`.
+
+Every example below assumes this setup (server-side, API key):
+
+```ts
+import { g8 } from "@graph8/sdk";
+
+g8.init({ apiKey: process.env.G8_API_KEY! });
+```
+
+Example values are placeholders that match each field's type. Only required fields are shown. Every example is type-checked against the SDK in CI.
+
+Exact TypeScript types for every operation on this page, and every schema they use: [documents.d.ts](documents.d.ts).
+
+## Operations
+
+| Call | HTTP | Tier | Scope | Summary |
+|---|---|---|---|---|
+| [`approveQuestionnaireAnswer`](#op-approvequestionnaireanswer) | `POST /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}/approve` | write | `companies:write` | Approve Questionnaire Answer |
+| [`createAccountCorrespondence`](#op-createaccountcorrespondence) | `POST /api/v1/accounts/{account_id}/correspondence` | write | `companies:write` | Create Account Correspondence |
+| [`createLibraryDocument`](#op-createlibrarydocument) | `POST /api/v1/library/documents` | write | `knowledge:write` | Create Library Document |
+| [`createLibraryFolder`](#op-createlibraryfolder) | `POST /api/v1/library/folders` | write | `knowledge:write` | Create Library Folder |
+| [`deleteAccountCorrespondence`](#op-deleteaccountcorrespondence) | `DELETE /api/v1/accounts/{account_id}/correspondence/{correspondence_id}` | destructive | `companies:delete` | Delete Account Correspondence |
+| [`deleteLibraryDocument`](#op-deletelibrarydocument) | `DELETE /api/v1/library/documents/{document_id}` | destructive | `knowledge:delete` | Delete Library Document |
+| [`deleteLibraryFolder`](#op-deletelibraryfolder) | `DELETE /api/v1/library/folders/{folder_id}` | destructive | `knowledge:delete` | Delete Library Folder |
+| [`deleteSecurityQuestionnaire`](#op-deletesecurityquestionnaire) | `DELETE /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}` | destructive | `companies:delete` | Delete Security Questionnaire |
+| [`downloadLibraryDocument`](#op-downloadlibrarydocument) | `GET /api/v1/library/documents/{document_id}/download` | read | `knowledge:read` | Download Library Document |
+| [`downloadSecurityQuestionnaire`](#op-downloadsecurityquestionnaire) | `GET /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/download` | read | `companies:read` | Download Security Questionnaire |
+| [`escalateQuestionnaireQuestion`](#op-escalatequestionnairequestion) | `POST /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}/escalate` | write | `companies:write` | Escalate Questionnaire Question |
+| [`flagQuestionnaireAnswer`](#op-flagquestionnaireanswer) | `POST /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}/mark-wrong` | write | `companies:write` | Flag Questionnaire Answer |
+| [`getLibraryDocument`](#op-getlibrarydocument) | `GET /api/v1/library/documents/{document_id}` | read | `knowledge:read` | Get Library Document |
+| [`getLibraryTree`](#op-getlibrarytree) | `GET /api/v1/library/tree` | read | `knowledge:read` | Get Library Tree |
+| [`getSecurityQuestionnaire`](#op-getsecurityquestionnaire) | `GET /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}` | read | `companies:read` | Get Security Questionnaire |
+| [`getSecurityQuestionnaireStatus`](#op-getsecurityquestionnairestatus) | `GET /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/status` | read | `companies:read` | Get Security Questionnaire Status |
+| [`importAccountCorrespondence`](#op-importaccountcorrespondence) | `POST /api/v1/accounts/{account_id}/correspondence/import` | write | `companies:write` | Import Account Correspondence |
+| [`listAccountCorrespondence`](#op-listaccountcorrespondence) | `GET /api/v1/accounts/{account_id}/correspondence` | read | `companies:read` | List Account Correspondence |
+| [`listLibraryDocuments`](#op-listlibrarydocuments) | `GET /api/v1/library/documents` | read | `knowledge:read` | List Library Documents |
+| [`listLibraryFolders`](#op-listlibraryfolders) | `GET /api/v1/library/folders` | read | `knowledge:read` | List Library Folders |
+| [`listSecurityQuestionnaires`](#op-listsecurityquestionnaires) | `GET /api/v1/accounts/{account_id}/security-questionnaires` | read | `companies:read` | List Security Questionnaires |
+| [`reextractAllLibraryDocuments`](#op-reextractalllibrarydocuments) | `POST /api/v1/library/documents/reextract-all` | write | `knowledge:write` | Reextract All Library Documents |
+| [`reextractLibraryDocument`](#op-reextractlibrarydocument) | `POST /api/v1/library/documents/{document_id}/reextract` | write | `knowledge:write` | Reextract Library Document |
+| [`refreshLibraryDocument`](#op-refreshlibrarydocument) | `POST /api/v1/library/documents/{document_id}/refresh` | write | `knowledge:write` | Refresh Library Document |
+| [`renameLibraryFolder`](#op-renamelibraryfolder) | `PUT /api/v1/library/folders/{folder_id}` | write | `knowledge:write` | Rename Library Folder |
+| [`restoreLibraryDocument`](#op-restorelibrarydocument) | `POST /api/v1/library/documents/{document_id}/restore` | write | `knowledge:write` | Restore Library Document |
+| [`saveQuestionnaireAnswerToLibrary`](#op-savequestionnaireanswertolibrary) | `POST /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}/save-to-library` | write | `companies:write` | Save Questionnaire Answer To Library |
+| [`skipQuestionnaireQuestion`](#op-skipquestionnairequestion) | `POST /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}/mark-skip` | write | `companies:write` | Skip Questionnaire Question |
+| [`updateLibraryDocument`](#op-updatelibrarydocument) | `PUT /api/v1/library/documents/{document_id}` | write | `knowledge:write` | Update Library Document |
+| [`updateQuestionnaireAnswer`](#op-updatequestionnaireanswer) | `PATCH /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}` | write | `companies:write` | Update Questionnaire Answer |
+| [`updateSecurityQuestionnaire`](#op-updatesecurityquestionnaire) | `PATCH /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}` | write | `companies:write` | Update Security Questionnaire |
+| [`uploadAccountCorrespondence`](#op-uploadaccountcorrespondence) | `POST /api/v1/accounts/{account_id}/correspondence/upload` | write | `companies:write` | Upload Account Correspondence |
+| [`uploadLibraryDocument`](#op-uploadlibrarydocument) | `POST /api/v1/library/documents/upload` | write | `knowledge:write` | Upload Library Document |
+| [`uploadLibraryDocuments`](#op-uploadlibrarydocuments) | `POST /api/v1/library/documents/upload-batch` | write | `knowledge:write` | Upload Library Documents |
+| [`uploadSecurityQuestionnaire`](#op-uploadsecurityquestionnaire) | `POST /api/v1/accounts/{account_id}/security-questionnaires` | billable | `knowledge:run` | Upload Security Questionnaire |
+
+---
+
+<a id="op-approvequestionnaireanswer"></a>
+
+### `g8.api.documents.approveQuestionnaireAnswer(input)`
+
+**Approve Questionnaire Answer**
+
+- HTTP: `POST /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}/approve`
+- operationId: `approve_questionnaire_answer_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__approve_post` (also callable as `g8.api.call("approve_questionnaire_answer_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__approve_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `companies:write`
+
+Confirm an answer as correct. No body - approving is the whole message.
+
+This is the signal that takes a questionnaire from `needs_review` towards
+done; it changes no text.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+| `path.question_id` | `string` | yes | Question id (UUID), from the questionnaire detail read. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.approveQuestionnaireAnswer({
+  path: { account_id: "id", questionnaire_id: "id", question_id: "id" },
+});
+```
+
+---
+
+<a id="op-createaccountcorrespondence"></a>
+
+### `g8.api.documents.createAccountCorrespondence(input)`
+
+**Create Account Correspondence**
+
+- HTTP: `POST /api/v1/accounts/{account_id}/correspondence`
+- operationId: `create_account_correspondence_accounts__account_id__correspondence_post` (also callable as `g8.api.call("create_account_correspondence_accounts__account_id__correspondence_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `companies:write`
+
+Record ONE message, call or note against an account.
+
+`correspondence_date` is when it HAPPENED, not when it was recorded - the
+generated documents read it as the relationship's timeline, so a wrong
+date is worse than a missing record.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `body` | `CorrespondenceCreateRequest` | yes | Request body; fields below. |
+
+Body fields (`CorrespondenceCreateRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.channel` | `string` | yes | `email`, `linkedin` or `call_notes`. |
+| `body.content` | `string` | yes | The body. Indexed for retrieval. |
+| `body.correspondence_date` | `string (date-time)` | yes | When it happened - NOT when it was recorded. Required. |
+| `body.direction` | `string` | yes | `inbound` or `outbound`. |
+| `body.external_id` | `string \| null` | no | Your id for this message. The app uses it to skip duplicates on `/correspondence/import`, so send it to make an import re-runnable. |
+| `body.stakeholder_id` | `string \| null` | no | The person at the account, when known. |
+| `body.subject` | `string \| null` | no |  |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.createAccountCorrespondence({
+  path: { account_id: "id" },
+  body: {
+    channel: "example",
+    content: "example",
+    correspondence_date: "2026-01-01T09:00:00Z",
+    direction: "example",
+  },
+});
+```
+
+---
+
+<a id="op-createlibrarydocument"></a>
+
+### `g8.api.documents.createLibraryDocument(input)`
+
+**Create Library Document**
+
+- HTTP: `POST /api/v1/library/documents`
+- operationId: `create_library_document_library_documents_post` (also callable as `g8.api.call("create_library_document_library_documents_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `knowledge:write`
+
+Create a TEXT document from a string.
+
+The counterpart to `/documents/upload`: no file, no extraction, the
+`content` you send is the content stored. Use it for material you already
+have as text.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body` | `LibraryDocumentCreateRequest` | yes | Request body; fields below. |
+
+Body fields (`LibraryDocumentCreateRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.content` | `string` | yes | The document text. Stored verbatim. |
+| `body.display_name` | `string` | yes | How the document is listed. |
+| `body.file_name` | `string` | yes | Stored file name. |
+| `body.folder_path` | `string` | yes | Where it lands, e.g. `Library/Case Studies`. The folder must already exist. |
+| `body.account_id` | `string \| null` | no | REQUIRED when `context_type` is `account`. |
+| `body.campaign_id` | `string \| null` | no | REQUIRED when `context_type` is `campaign`. |
+| `body.context_type` | `"global" \| "campaign" \| "account"` | no | Whose library this belongs to. default `"global"` |
+| `body.copilot_session_id` | `string \| null` | no | The Copilot session that produced it, when `source` is `copilot`. |
+| `body.source` | `string` | no | Where the text came from; `manual` or `copilot` are what the app writes. NOT validated by the app, so any value is accepted and stored as given. default `"manual"` |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.createLibraryDocument({
+  body: {
+    content: "example",
+    display_name: "Example",
+    file_name: "Example",
+    folder_path: "example",
+  },
+});
+```
+
+---
+
+<a id="op-createlibraryfolder"></a>
+
+### `g8.api.documents.createLibraryFolder(input)`
+
+**Create Library Folder**
+
+- HTTP: `POST /api/v1/library/folders`
+- operationId: `create_library_folder_library_folders_post` (also callable as `g8.api.call("create_library_folder_library_folders_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `knowledge:write`
+
+Create a folder. `parent_path` `Library` is the root.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body` | `LibraryFolderCreateRequest` | yes | Request body; fields below. |
+
+Body fields (`LibraryFolderCreateRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.name` | `string` | yes | Folder name, not a path. |
+| `body.account_id` | `string \| null` | no | REQUIRED when `context_type` is `account`. |
+| `body.campaign_id` | `string \| null` | no | REQUIRED when `context_type` is `campaign`. |
+| `body.context_type` | `"global" \| "campaign" \| "account"` | no | default `"global"` |
+| `body.parent_path` | `string` | no | Where it hangs. `Library` is the root. default `"Library"` |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.createLibraryFolder({ body: { name: "Example" } });
+```
+
+---
+
+<a id="op-deleteaccountcorrespondence"></a>
+
+### `g8.api.documents.deleteAccountCorrespondence(input)`
+
+**Delete Account Correspondence**
+
+- HTTP: `DELETE /api/v1/accounts/{account_id}/correspondence/{correspondence_id}`
+- operationId: `delete_account_correspondence_accounts__account_id__correspondence__correspondence_id__delete` (also callable as `g8.api.call("delete_account_correspondence_accounts__account_id__correspondence__correspondence_id__delete", input)`)
+- Tier: `destructive`: deletes or archives data; confirm with a human before calling
+- API-key scope: `companies:delete`
+
+Delete one correspondence record. 204, matching the app.
+
+The generated documents that already cited it are not rewritten - this
+removes the source, not its consequences.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.correspondence_id` | `string` | yes | The record to remove. |
+
+**Returns**: no JSON body is declared for the success response.
+
+**Example**
+
+```ts
+// destructive: deletes or archives data; confirm with a human before calling.
+const result = await g8.api.documents.deleteAccountCorrespondence({ path: { account_id: "id", correspondence_id: "id" } });
+```
+
+---
+
+<a id="op-deletelibrarydocument"></a>
+
+### `g8.api.documents.deleteLibraryDocument(input)`
+
+**Delete Library Document**
+
+- HTTP: `DELETE /api/v1/library/documents/{document_id}`
+- operationId: `delete_library_document_library_documents__document_id__delete` (also callable as `g8.api.call("delete_library_document_library_documents__document_id__delete", input)`)
+- Tier: `destructive`: deletes or archives data; confirm with a human before calling
+- API-key scope: `knowledge:delete`
+
+Delete a document PERMANENTLY. The row and its bytes go.
+
+This is the app's hard delete, not the soft delete `/documents/{id}/restore`
+reverses - a document removed here cannot be restored.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.document_id` | `string` | yes | Document id. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+// destructive: deletes or archives data; confirm with a human before calling.
+const result = await g8.api.documents.deleteLibraryDocument({ path: { document_id: "id" } });
+```
+
+---
+
+<a id="op-deletelibraryfolder"></a>
+
+### `g8.api.documents.deleteLibraryFolder(input)`
+
+**Delete Library Folder**
+
+- HTTP: `DELETE /api/v1/library/folders/{folder_id}`
+- operationId: `delete_library_folder_library_folders__folder_id__delete` (also callable as `g8.api.call("delete_library_folder_library_folders__folder_id__delete", input)`)
+- Tier: `destructive`: deletes or archives data; confirm with a human before calling
+- API-key scope: `knowledge:delete`
+
+Delete a folder. It must be EMPTY - this never deletes documents.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.folder_id` | `string` | yes | Folder id. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+// destructive: deletes or archives data; confirm with a human before calling.
+const result = await g8.api.documents.deleteLibraryFolder({ path: { folder_id: "id" } });
+```
+
+---
+
+<a id="op-deletesecurityquestionnaire"></a>
+
+### `g8.api.documents.deleteSecurityQuestionnaire(input)`
+
+**Delete Security Questionnaire**
+
+- HTTP: `DELETE /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}`
+- operationId: `delete_security_questionnaire_accounts__account_id__security_questionnaires__questionnaire_id__delete` (also callable as `g8.api.call("delete_security_questionnaire_accounts__account_id__security_questionnaires__questionnaire_id__delete", input)`)
+- Tier: `destructive`: deletes or archives data; confirm with a human before calling
+- API-key scope: `companies:delete`
+
+Delete a questionnaire AND every question on it. Not reversible.
+
+204 with no body, matching the app - there is nothing left to describe.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+
+**Returns**: no JSON body is declared for the success response.
+
+**Example**
+
+```ts
+// destructive: deletes or archives data; confirm with a human before calling.
+const result = await g8.api.documents.deleteSecurityQuestionnaire({ path: { account_id: "id", questionnaire_id: "id" } });
+```
+
+---
+
+<a id="op-downloadlibrarydocument"></a>
+
+### `g8.api.documents.downloadLibraryDocument(input)`
+
+**Download Library Document**
+
+- HTTP: `GET /api/v1/library/documents/{document_id}/download`
+- operationId: `download_library_document_library_documents__document_id__download_get` (also callable as `g8.api.call("download_library_document_library_documents__document_id__download_get", input)`)
+- Tier: `read`: reads data; safe to call and to retry
+- API-key scope: `knowledge:read`
+
+The document's ORIGINAL bytes, with its own content type.
+
+The one route on this surface that is NOT the `ApiResponse` envelope: a
+PDF wrapped in JSON is not a PDF. Same precedent as
+`GET /knowledge/export`. Content type and filename come from the stored
+document, so a text document comes back as text and a PDF as a PDF.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.document_id` | `string` | yes | Document id. |
+
+**Returns**: no JSON body is declared for the success response.
+
+**Example**
+
+```ts
+const result = await g8.api.documents.downloadLibraryDocument({ path: { document_id: "id" } });
+```
+
+---
+
+<a id="op-downloadsecurityquestionnaire"></a>
+
+### `g8.api.documents.downloadSecurityQuestionnaire(input)`
+
+**Download Security Questionnaire**
+
+- HTTP: `GET /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/download`
+- operationId: `download_security_questionnaire_accounts__account_id__security_questionnaires__questionnaire_id__download_get` (also callable as `g8.api.call("download_security_questionnaire_accounts__account_id__security_questionnaires__questionnaire_id__download_get", input)`)
+- Tier: `read`: reads data; safe to call and to retry
+- API-key scope: `companies:read`
+
+The FILLED workbook, as XLSX bytes.
+
+Not the `ApiResponse` envelope - a spreadsheet wrapped in JSON is not a
+spreadsheet. Same precedent as `GET /knowledge/export`.
+
+Only available once the questionnaire is `completed` or `needs_review`;
+asking earlier is refused rather than returning a half-filled file.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+
+**Returns**: no JSON body is declared for the success response.
+
+**Example**
+
+```ts
+const result = await g8.api.documents.downloadSecurityQuestionnaire({ path: { account_id: "id", questionnaire_id: "id" } });
+```
+
+---
+
+<a id="op-escalatequestionnairequestion"></a>
+
+### `g8.api.documents.escalateQuestionnaireQuestion(input)`
+
+**Escalate Questionnaire Question**
+
+- HTTP: `POST /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}/escalate`
+- operationId: `escalate_questionnaire_question_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__escalate_post` (also callable as `g8.api.call("escalate_questionnaire_question_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__escalate_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `companies:write`
+
+Hand a question to the security team.
+
+For the questions the answer library genuinely cannot settle. `notes` is
+what the person picking it up reads, so it is worth sending even though
+the app allows it to be empty.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+| `path.question_id` | `string` | yes | Question id (UUID), from the questionnaire detail read. |
+| `body` | `QuestionNoteRequest` | no | Request body; fields below. |
+
+Body fields (`QuestionNoteRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.notes` | `string \| null` | no | Why. Shown to whoever picks the question up. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.escalateQuestionnaireQuestion({
+  path: { account_id: "id", questionnaire_id: "id", question_id: "id" },
+});
+```
+
+---
+
+<a id="op-flagquestionnaireanswer"></a>
+
+### `g8.api.documents.flagQuestionnaireAnswer(input)`
+
+**Flag Questionnaire Answer**
+
+- HTTP: `POST /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}/mark-wrong`
+- operationId: `flag_questionnaire_answer_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__mark_wrong_post` (also callable as `g8.api.call("flag_questionnaire_answer_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__mark_wrong_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `companies:write`
+
+Say the AI's answer is WRONG and needs replacing.
+
+The feedback signal, not a fix: it flags the answer for a human to
+rewrite with `PATCH .../questions/{question_id}`.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+| `path.question_id` | `string` | yes | Question id (UUID), from the questionnaire detail read. |
+| `body` | `QuestionNoteRequest` | no | Request body; fields below. |
+
+Body fields (`QuestionNoteRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.notes` | `string \| null` | no | Why. Shown to whoever picks the question up. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.flagQuestionnaireAnswer({
+  path: { account_id: "id", questionnaire_id: "id", question_id: "id" },
+});
+```
+
+---
+
+<a id="op-getlibrarydocument"></a>
+
+### `g8.api.documents.getLibraryDocument(input)`
+
+**Get Library Document**
+
+- HTTP: `GET /api/v1/library/documents/{document_id}`
+- operationId: `get_library_document_library_documents__document_id__get` (also callable as `g8.api.call("get_library_document_library_documents__document_id__get", input)`)
+- Tier: `read`: reads data; safe to call and to retry
+- API-key scope: `knowledge:read`
+
+One document, with its extracted text.
+
+For the original bytes use `/documents/{id}/download`.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.document_id` | `string` | yes | Document id. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.getLibraryDocument({ path: { document_id: "id" } });
+```
+
+---
+
+<a id="op-getlibrarytree"></a>
+
+### `g8.api.documents.getLibraryTree(input?)`
+
+**Get Library Tree**
+
+- HTTP: `GET /api/v1/library/tree`
+- operationId: `get_library_tree_library_tree_get` (also callable as `g8.api.call("get_library_tree_library_tree_get", input)`)
+- Tier: `read`: reads data; safe to call and to retry
+- API-key scope: `knowledge:read`
+
+Folders and their documents, nested - the whole library in one call.
+
+The read to build a file browser from. `/library/folders` and
+`/library/documents` are the flat halves of the same data.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `query.context_type` | `"global" \| "campaign" \| "account"` | no | Whose library to read. `campaign` and `account` additionally need the matching id below; without it the app reads the GLOBAL library, which is a different set of documents rather than an error. default `"global"` |
+| `query.campaign_id` | `string \| null` | no | Campaign id. Only read when `context_type` is `campaign`. |
+| `query.account_id` | `string \| null` | no | Account id. Only read when `context_type` is `account`. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.getLibraryTree();
+```
+
+---
+
+<a id="op-getsecurityquestionnaire"></a>
+
+### `g8.api.documents.getSecurityQuestionnaire(input)`
+
+**Get Security Questionnaire**
+
+- HTTP: `GET /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}`
+- operationId: `get_security_questionnaire_accounts__account_id__security_questionnaires__questionnaire_id__get` (also callable as `g8.api.call("get_security_questionnaire_accounts__account_id__security_questionnaires__questionnaire_id__get", input)`)
+- Tier: `read`: reads data; safe to call and to retry
+- API-key scope: `companies:read`
+
+One questionnaire, with every question and its answer.
+
+The read the review loop is built on: each question carries the AI's
+answer, the human's edit if there is one, the confidence, and whether it
+has been approved, escalated, skipped or marked wrong.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.getSecurityQuestionnaire({ path: { account_id: "id", questionnaire_id: "id" } });
+```
+
+---
+
+<a id="op-getsecurityquestionnairestatus"></a>
+
+### `g8.api.documents.getSecurityQuestionnaireStatus(input)`
+
+**Get Security Questionnaire Status**
+
+- HTTP: `GET /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/status`
+- operationId: `get_security_questionnaire_status_accounts__account_id__security_questionnaires__questionnaire_id__status_get` (also callable as `g8.api.call("get_security_questionnaire_status_accounts__account_id__security_questionnaires__questionnaire_id__status_get", input)`)
+- Tier: `read`: reads data; safe to call and to retry
+- API-key scope: `companies:read`
+
+Progress only - the cheap read to POLL.
+
+The full detail read returns every question and grows with the document;
+this one answers "is it done yet" without transferring it.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.getSecurityQuestionnaireStatus({ path: { account_id: "id", questionnaire_id: "id" } });
+```
+
+---
+
+<a id="op-importaccountcorrespondence"></a>
+
+### `g8.api.documents.importAccountCorrespondence(input)`
+
+**Import Account Correspondence**
+
+- HTTP: `POST /api/v1/accounts/{account_id}/correspondence/import`
+- operationId: `import_account_correspondence_accounts__account_id__correspondence_import_post` (also callable as `g8.api.call("import_account_correspondence_accounts__account_id__correspondence_import_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `companies:write`
+
+Import a batch of correspondence. PARTIAL SUCCESS IS NORMAL.
+
+The response reports `imported`, `skipped` and `errors` and the status
+stays 200 even when rows failed - read the body, not the status. Send
+`external_id` on each item to make the import re-runnable: a row whose
+`external_id` is already stored is SKIPPED rather than duplicated.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `body` | `CorrespondenceImportRequest` | yes | Request body; fields below. |
+
+Body fields (`CorrespondenceImportRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.items` | `Array<CorrespondenceCreateRequest>` | yes | The records. Bounded at 500 by this contract - the app has no ceiling, and an unbounded body on a synchronous per-row insert is a request that times out halfway through having written some rows. min items 1, max items 500 |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.importAccountCorrespondence({
+  path: { account_id: "id" },
+  body: {
+    items: [
+      {
+        channel: "example",
+        content: "example",
+        correspondence_date: "2026-01-01T09:00:00Z",
+        direction: "example",
+      },
+    ],
+  },
+});
+```
+
+---
+
+<a id="op-listaccountcorrespondence"></a>
+
+### `g8.api.documents.listAccountCorrespondence(input)`
+
+**List Account Correspondence**
+
+- HTTP: `GET /api/v1/accounts/{account_id}/correspondence`
+- operationId: `list_account_correspondence_accounts__account_id__correspondence_get` (also callable as `g8.api.call("list_account_correspondence_accounts__account_id__correspondence_get", input)`)
+- Tier: `read`: reads data; safe to call and to retry
+- API-key scope: `companies:read`
+
+What has been said to this account, across channels.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `query.channel` | `string \| null` | no | `email`, `linkedin` or `call_notes`. Omit for all. |
+| `query.stakeholder_id` | `string \| null` | no | Only what was said to or by this person. |
+| `query.limit` | `integer` | no | default `50` |
+| `query.offset` | `integer` | no | default `0` |
+
+**Returns** `ApiResponse_DocumentListPayload_`: the envelope `{ data: DocumentListPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+`DocumentListPayload` fields:
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `items` | `Array<unknown>` | no |  |
+
+**Example**
+
+```ts
+const result = await g8.api.documents.listAccountCorrespondence({ path: { account_id: "id" } });
+```
+
+---
+
+<a id="op-listlibrarydocuments"></a>
+
+### `g8.api.documents.listLibraryDocuments(input?)`
+
+**List Library Documents**
+
+- HTTP: `GET /api/v1/library/documents`
+- operationId: `list_library_documents_library_documents_get` (also callable as `g8.api.call("list_library_documents_library_documents_get", input)`)
+- Tier: `read`: reads data; safe to call and to retry
+- API-key scope: `knowledge:read`
+
+What is in the library.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `query.context_type` | `"global" \| "campaign" \| "account"` | no | Whose library to read. `campaign` and `account` additionally need the matching id below; without it the app reads the GLOBAL library, which is a different set of documents rather than an error. default `"global"` |
+| `query.campaign_id` | `string \| null` | no | Campaign id. Only read when `context_type` is `campaign`. |
+| `query.account_id` | `string \| null` | no | Account id. Only read when `context_type` is `account`. |
+| `query.folder_path` | `string \| null` | no | Only this folder. |
+| `query.include_deleted` | `boolean` | no | Include soft-deleted documents, which is how you find something to pass to `/documents/{id}/restore`. default `false` |
+
+**Returns** `ApiResponse_DocumentListPayload_`: the envelope `{ data: DocumentListPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+`DocumentListPayload` fields:
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `items` | `Array<unknown>` | no |  |
+
+**Example**
+
+```ts
+const result = await g8.api.documents.listLibraryDocuments();
+```
+
+---
+
+<a id="op-listlibraryfolders"></a>
+
+### `g8.api.documents.listLibraryFolders(input?)`
+
+**List Library Folders**
+
+- HTTP: `GET /api/v1/library/folders`
+- operationId: `list_library_folders_library_folders_get` (also callable as `g8.api.call("list_library_folders_library_folders_get", input)`)
+- Tier: `read`: reads data; safe to call and to retry
+- API-key scope: `knowledge:read`
+
+The folders, flat. For the nested shape use `/library/tree`.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `query.context_type` | `"global" \| "campaign" \| "account"` | no | Whose library to read. `campaign` and `account` additionally need the matching id below; without it the app reads the GLOBAL library, which is a different set of documents rather than an error. default `"global"` |
+| `query.campaign_id` | `string \| null` | no | Campaign id. Only read when `context_type` is `campaign`. |
+| `query.account_id` | `string \| null` | no | Account id. Only read when `context_type` is `account`. |
+
+**Returns** `ApiResponse_DocumentListPayload_`: the envelope `{ data: DocumentListPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+`DocumentListPayload` fields:
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `items` | `Array<unknown>` | no |  |
+
+**Example**
+
+```ts
+const result = await g8.api.documents.listLibraryFolders();
+```
+
+---
+
+<a id="op-listsecurityquestionnaires"></a>
+
+### `g8.api.documents.listSecurityQuestionnaires(input)`
+
+**List Security Questionnaires**
+
+- HTTP: `GET /api/v1/accounts/{account_id}/security-questionnaires`
+- operationId: `list_security_questionnaires_accounts__account_id__security_questionnaires_get` (also callable as `g8.api.call("list_security_questionnaires_accounts__account_id__security_questionnaires_get", input)`)
+- Tier: `read`: reads data; safe to call and to retry
+- API-key scope: `companies:read`
+
+Every questionnaire for this account.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `query.status` | `string \| null` | no | Only questionnaires in this state: `processing`, `completed`, `needs_review` or `failed`. Omit for all. |
+| `query.limit` | `integer` | no | default `50` |
+| `query.offset` | `integer` | no | default `0` |
+
+**Returns** `ApiResponse_DocumentListPayload_`: the envelope `{ data: DocumentListPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+`DocumentListPayload` fields:
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `items` | `Array<unknown>` | no |  |
+
+**Example**
+
+```ts
+const result = await g8.api.documents.listSecurityQuestionnaires({ path: { account_id: "id" } });
+```
+
+---
+
+<a id="op-reextractalllibrarydocuments"></a>
+
+### `g8.api.documents.reextractAllLibraryDocuments(input?)`
+
+**Reextract All Library Documents**
+
+- HTTP: `POST /api/v1/library/documents/reextract-all`
+- operationId: `reextract_all_library_documents_library_documents_reextract_all_post` (also callable as `g8.api.call("reextract_all_library_documents_library_documents_reextract_all_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `knowledge:write`
+
+Re-run text extraction across a folder, or the whole library.
+
+Use it after an extraction improvement, or when documents uploaded before
+a fix have empty or garbled text. It replaces the stored text; the
+original bytes are untouched.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `query.folder_path` | `string \| null` | no | Limit to one folder. Omit to re-extract EVERY PDF and DOCX in the library, which on a large library is a long job. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.reextractAllLibraryDocuments();
+```
+
+---
+
+<a id="op-reextractlibrarydocument"></a>
+
+### `g8.api.documents.reextractLibraryDocument(input)`
+
+**Reextract Library Document**
+
+- HTTP: `POST /api/v1/library/documents/{document_id}/reextract`
+- operationId: `reextract_library_document_library_documents__document_id__reextract_post` (also callable as `g8.api.call("reextract_library_document_library_documents__document_id__reextract_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `knowledge:write`
+
+Re-run text extraction on ONE PDF or DOCX.
+
+The single-document form of `/documents/reextract-all`. The stored bytes
+are the input and are not changed.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.document_id` | `string` | yes | Document id. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.reextractLibraryDocument({ path: { document_id: "id" } });
+```
+
+---
+
+<a id="op-refreshlibrarydocument"></a>
+
+### `g8.api.documents.refreshLibraryDocument(input)`
+
+**Refresh Library Document**
+
+- HTTP: `POST /api/v1/library/documents/{document_id}/refresh`
+- operationId: `refresh_library_document_library_documents__document_id__refresh_post` (also callable as `g8.api.call("refresh_library_document_library_documents__document_id__refresh_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `knowledge:write`
+
+Re-fetch a document that was SCRAPED from a URL.
+
+Only meaningful for a document whose `source` is `scrape`; anything else
+is refused, because there is no URL to go back to.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.document_id` | `string` | yes | Document id. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.refreshLibraryDocument({ path: { document_id: "id" } });
+```
+
+---
+
+<a id="op-renamelibraryfolder"></a>
+
+### `g8.api.documents.renameLibraryFolder(input)`
+
+**Rename Library Folder**
+
+- HTTP: `PUT /api/v1/library/folders/{folder_id}`
+- operationId: `rename_library_folder_library_folders__folder_id__put` (also callable as `g8.api.call("rename_library_folder_library_folders__folder_id__put", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `knowledge:write`
+
+Rename a folder. The documents inside keep their ids.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.folder_id` | `string` | yes | Folder id. |
+| `body` | `LibraryFolderRenameRequest` | yes | Request body; fields below. |
+
+Body fields (`LibraryFolderRenameRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.name` | `string` | yes | The new name. Documents inside keep their ids. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.renameLibraryFolder({ path: { folder_id: "id" }, body: { name: "Example" } });
+```
+
+---
+
+<a id="op-restorelibrarydocument"></a>
+
+### `g8.api.documents.restoreLibraryDocument(input)`
+
+**Restore Library Document**
+
+- HTTP: `POST /api/v1/library/documents/{document_id}/restore`
+- operationId: `restore_library_document_library_documents__document_id__restore_post` (also callable as `g8.api.call("restore_library_document_library_documents__document_id__restore_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `knowledge:write`
+
+Undo a SOFT delete.
+
+Find the id with `GET /library/documents?include_deleted=true`. This
+cannot bring back a document removed by `DELETE`, which is permanent.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.document_id` | `string` | yes | Document id. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.restoreLibraryDocument({ path: { document_id: "id" } });
+```
+
+---
+
+<a id="op-savequestionnaireanswertolibrary"></a>
+
+### `g8.api.documents.saveQuestionnaireAnswerToLibrary(input)`
+
+**Save Questionnaire Answer To Library**
+
+- HTTP: `POST /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}/save-to-library`
+- operationId: `save_questionnaire_answer_to_library_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__save_to_library_post` (also callable as `g8.api.call("save_questionnaire_answer_to_library_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__save_to_library_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `companies:write`
+
+Promote this answer into the reusable answer library.
+
+The compounding step: an answer saved here is what the NEXT questionnaire
+is answered from, so a reviewed answer stops being a one-off edit and
+starts being org knowledge.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+| `path.question_id` | `string` | yes | Question id (UUID), from the questionnaire detail read. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.saveQuestionnaireAnswerToLibrary({
+  path: { account_id: "id", questionnaire_id: "id", question_id: "id" },
+});
+```
+
+---
+
+<a id="op-skipquestionnairequestion"></a>
+
+### `g8.api.documents.skipQuestionnaireQuestion(input)`
+
+**Skip Questionnaire Question**
+
+- HTTP: `POST /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}/mark-skip`
+- operationId: `skip_questionnaire_question_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__mark_skip_post` (also callable as `g8.api.call("skip_questionnaire_question_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__mark_skip_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `companies:write`
+
+Mark a question as NOT APPLICABLE to this organisation.
+
+Different from escalating: nobody is going to answer this one, and the
+filled workbook says so rather than leaving it blank.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+| `path.question_id` | `string` | yes | Question id (UUID), from the questionnaire detail read. |
+| `body` | `QuestionNoteRequest` | no | Request body; fields below. |
+
+Body fields (`QuestionNoteRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.notes` | `string \| null` | no | Why. Shown to whoever picks the question up. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.skipQuestionnaireQuestion({
+  path: { account_id: "id", questionnaire_id: "id", question_id: "id" },
+});
+```
+
+---
+
+<a id="op-updatelibrarydocument"></a>
+
+### `g8.api.documents.updateLibraryDocument(input)`
+
+**Update Library Document**
+
+- HTTP: `PUT /api/v1/library/documents/{document_id}`
+- operationId: `update_library_document_library_documents__document_id__put` (also callable as `g8.api.call("update_library_document_library_documents__document_id__put", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `knowledge:write`
+
+Edit a document: rename it, replace its text, or move it.
+
+PARTIAL despite the PUT verb - an omitted field is left alone, which is
+the app's own shape. `meta_data` REPLACES the stored object rather than
+merging into it.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.document_id` | `string` | yes | Document id. |
+| `body` | `LibraryDocumentUpdateRequest` | yes | Request body; fields below. |
+
+Body fields (`LibraryDocumentUpdateRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.content` | `string \| null` | no | Replaces the whole document body when given. |
+| `body.display_name` | `string \| null` | no |  |
+| `body.folder_path` | `string \| null` | no | Moves the document to another folder when given. |
+| `body.meta_data` | `Record<string, unknown> \| null` | no | Free-form metadata. REPLACES the stored object rather than merging. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.updateLibraryDocument({ path: { document_id: "id" }, body: {} });
+```
+
+---
+
+<a id="op-updatequestionnaireanswer"></a>
+
+### `g8.api.documents.updateQuestionnaireAnswer(input)`
+
+**Update Questionnaire Answer**
+
+- HTTP: `PATCH /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}/questions/{question_id}`
+- operationId: `update_questionnaire_answer_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__patch` (also callable as `g8.api.call("update_questionnaire_answer_accounts__account_id__security_questionnaires__questionnaire_id__questions__question_id__patch", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `companies:write`
+
+Replace one answer with a human-written one.
+
+The edit lands in `final_answer` and is what the downloaded workbook
+carries. The model's original answer is KEPT alongside it, so the two can
+still be compared afterwards.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+| `path.question_id` | `string` | yes | Question id (UUID), from the questionnaire detail read. |
+| `body` | `QuestionAnswerUpdateRequest` | yes | Request body; fields below. |
+
+Body fields (`QuestionAnswerUpdateRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.final_answer` | `string` | yes | The edited answer text. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.updateQuestionnaireAnswer({
+  path: { account_id: "id", questionnaire_id: "id", question_id: "id" },
+  body: { final_answer: "example" },
+});
+```
+
+---
+
+<a id="op-updatesecurityquestionnaire"></a>
+
+### `g8.api.documents.updateSecurityQuestionnaire(input)`
+
+**Update Security Questionnaire**
+
+- HTTP: `PATCH /api/v1/accounts/{account_id}/security-questionnaires/{questionnaire_id}`
+- operationId: `update_security_questionnaire_accounts__account_id__security_questionnaires__questionnaire_id__patch` (also callable as `g8.api.call("update_security_questionnaire_accounts__account_id__security_questionnaires__questionnaire_id__patch", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `companies:write`
+
+Rename a questionnaire. Metadata only - it never touches an answer.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+| `path.questionnaire_id` | `string` | yes | Questionnaire id (UUID). |
+| `body` | `QuestionnaireUpdateRequest` | yes | Request body; fields below. |
+
+Body fields (`QuestionnaireUpdateRequest`):
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `body.display_name` | `string \| null` | no | User-friendly name for the questionnaire. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.updateSecurityQuestionnaire({ path: { account_id: "id", questionnaire_id: "id" }, body: {} });
+```
+
+---
+
+<a id="op-uploadaccountcorrespondence"></a>
+
+### `g8.api.documents.uploadAccountCorrespondence(input)`
+
+**Upload Account Correspondence**
+
+- HTTP: `POST /api/v1/accounts/{account_id}/correspondence/upload`
+- operationId: `upload_account_correspondence_accounts__account_id__correspondence_upload_post` (also callable as `g8.api.call("upload_account_correspondence_accounts__account_id__correspondence_upload_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `companies:write`
+
+Record correspondence from a FILE rather than a JSON body.
+
+The route for a forwarded `.eml` or an exported thread: the file's text
+becomes the content, and the record is created in the same call.
+
+This handler is `async def` but takes the SYNC session, so it is awaited
+directly rather than going through the threadpool helper.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.uploadAccountCorrespondence({ path: { account_id: "id" } });
+```
+
+---
+
+<a id="op-uploadlibrarydocument"></a>
+
+### `g8.api.documents.uploadLibraryDocument(input?)`
+
+**Upload Library Document**
+
+- HTTP: `POST /api/v1/library/documents/upload`
+- operationId: `upload_library_document_library_documents_upload_post` (also callable as `g8.api.call("upload_library_document_library_documents_upload_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `knowledge:write`
+
+Upload ONE file to the Library. Multipart, like the app.
+
+The app's rules are enforced by the app and documented here so a caller
+sees them before spending the upload: 50 MB maximum, and the extension
+must be one of pdf, docx, doc, txt, md, mdx, csv, json, png, jpg, jpeg,
+xls, xlsx, ppt, pptx. Text-bearing types are extracted and indexed for
+retrieval; the rest are stored as bytes.
+
+**Input**: none.
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+const result = await g8.api.documents.uploadLibraryDocument();
+```
+
+---
+
+<a id="op-uploadlibrarydocuments"></a>
+
+### `g8.api.documents.uploadLibraryDocuments(input?)`
+
+**Upload Library Documents**
+
+- HTTP: `POST /api/v1/library/documents/upload-batch`
+- operationId: `upload_library_documents_library_documents_upload_batch_post` (also callable as `g8.api.call("upload_library_documents_library_documents_upload_batch_post", input)`)
+- Tier: `write`: creates or changes records; the SDK sends an Idempotency-Key and retries transient failures
+- API-key scope: `knowledge:write`
+
+Upload SEVERAL files in one request.
+
+PARTIAL SUCCESS IS THE NORMAL OUTCOME. Each file is processed on its own
+session, so one rejected file does not roll the others back and the status
+stays 201. Read the per-file result in the body, not the HTTP status.
+
+Named `upload-batch` rather than the app's `upload-multi`: the same
+operation, spelled the way the rest of this contract spells a bulk verb.
+
+**Input**: none.
+
+**Returns** `ApiResponse_DocumentListPayload_`: the envelope `{ data: DocumentListPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+`DocumentListPayload` fields:
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `items` | `Array<unknown>` | no |  |
+
+**Example**
+
+```ts
+const result = await g8.api.documents.uploadLibraryDocuments();
+```
+
+---
+
+<a id="op-uploadsecurityquestionnaire"></a>
+
+### `g8.api.documents.uploadSecurityQuestionnaire(input)`
+
+**Upload Security Questionnaire**
+
+- HTTP: `POST /api/v1/accounts/{account_id}/security-questionnaires`
+- operationId: `upload_security_questionnaire_accounts__account_id__security_questionnaires_post` (also callable as `g8.api.call("upload_security_questionnaire_accounts__account_id__security_questionnaires_post", input)`)
+- Tier: `billable`: spends credits; the SDK never blindly retries it; confirm cost before bulk use
+- API-key scope: `knowledge:run`
+
+Upload a questionnaire and start answering it. ASYNCHRONOUS.
+
+Answering runs in the background against the org's answer library and
+policy documents, so this returns as soon as the file is accepted. Poll
+`/security-questionnaires/{id}/status` for progress and
+`/security-questionnaires/{id}` for the questions as they land.
+
+Multipart, like the app.
+
+**Input**
+
+| Field | Type | Required | Description |
+|---|---|---|---|
+| `path.account_id` | `string` | yes | The account the questionnaire came from. |
+
+**Returns** `ApiResponse_DocumentPayload_`: the envelope `{ data: DocumentPayload, ... }`. `g8.api` returns the envelope as-is, so read `result.data`.
+
+Other envelope fields: `pagination` (PaginationMeta).
+
+**Example**
+
+```ts
+// billable: spends credits; the SDK never blindly retries it; confirm cost before bulk use.
+const result = await g8.api.documents.uploadSecurityQuestionnaire({ path: { account_id: "id" } });
+```