My Data — documents

Uploading, listing, organising, sharing and deleting the files in My Data, plus the AI features that run over them: summaries, generated questions, semantic search and document Q&A.

Three stores, one shape

Files are kept in module-specific stores that share the same URL shape:

Store	Prefix	Holds
Semantic search	`/semantic-search/`	PDFs, Office documents, text, images — chunked and embedded for retrieval
Database search	`/database-search/`	CSV/Excel treated as tabular data for the database chat
Spreadsheet search	`/spreadsheet-search/`	Spreadsheets for the spreadsheet chat

Each store exposes upload/, documents/, <uid>/delete/, folder/… and a chat/ endpoint. The semantic-search store is the richest and is the one the My Data page is built on; the other two are documented where they differ.

Processing status

An upload returns immediately; embedding happens in the background. Poll GET /semantic-search/documents/status/ (batched) or read the status field of GET /semantic-search/documents/ until the document is completed.

Endpoints

Method	Path	Purpose
POST	`/semantic-search/upload/`	Upload a document
POST	`/semantic-search/upload/batch/`	Upload several documents at once
GET	`/semantic-search/upload/batch/{batch_id}/status/`	Status of a batch upload
GET	`/semantic-search/documents/`	List my documents
GET	`/semantic-search/documents/status/`	Documents still processing
POST	`/semantic-search/documents/status/`	Status of specific documents
GET	`/semantic-search/documents/{uid}/`	Get one document
PUT	`/semantic-search/{uid}/update/`	Move a document to a folder
DELETE	`/semantic-search/{uid}/delete/`	Move a document to the trash
GET	`/semantic-search/documents/trash/`	List trashed documents
POST	`/semantic-search/{uid}/restore/`	Restore a document from the trash
DELETE	`/semantic-search/{uid}/permanent-delete/`	Delete a document permanently
DELETE	`/semantic-search/folder/delete/`	Move a folder to the trash
POST	`/semantic-search/folder/move/`	Move or rename a folder
GET	`/semantic-search/documents/{uid}/users/`	Who a document is shared with
POST	`/semantic-search/documents/{uid}/users/add/`	Share a document with users
POST	`/semantic-search/documents/{uid}/users/remove/`	Stop sharing a document with users
GET	`/semantic-search/documents/requests/`	Access requests
POST	`/semantic-search/documents/request/`	Request access to a document
DELETE	`/semantic-search/documents/request/`	Cancel my pending request
POST	`/semantic-search/documents/approve/`	Approve an access request
POST	`/semantic-search/documents/deny/`	Deny an access request
POST	`/semantic-search/documents/revoke/`	Revoke previously granted access
GET	`/semantic-search/{uid}/summary/`	Document summary & suggested prompts
POST	`/semantic-search/{uid}/summary/generate/`	Generate the summary & prompts
GET	`/semantic-search/{uid}/questions/`	Suggested prompts for a document
POST	`/semantic-search/questions/`	Add suggested prompts to a document
POST	`/semantic-search/summaries/`	Template-generated summaries
POST	`/semantic-search/rewrite-summary/`	Generate a summary from a prompt template
POST	`/semantic-search/search/`	Semantic search across my documents
POST	`/semantic-search/re-embed/`	Re-index documents
POST	`/database-search/upload/`	Upload a CSV/Excel for the database chat
GET	`/database-search/documents/`	List my database-search files
PUT	`/database-search/{uid}/update/`	Move a database-search file to a folder
DELETE	`/database-search/{uid}/delete/`	Delete a database-search file
DELETE	`/database-search/folder/delete/`	Delete a database-search folder
POST	`/database-search/folder/move/`	Move or rename a database-search folder
POST	`/database-search/chat/`	Ask a question about a CSV/Excel file
POST	`/database-search/store/`	Store a database-chat message
GET	`/database-search/{uid}/thoughts/`	Analysis notes for a database-search file
GET	`/database-search/`	All my database-chat messages
POST	`/spreadsheet-search/upload/`	Upload a spreadsheet for the spreadsheet chat
GET	`/spreadsheet-search/documents/`	List my spreadsheet-search files
DELETE	`/spreadsheet-search/{uid}/delete/`	Delete a spreadsheet-search file
POST	`/spreadsheet-search/chat/`	Chat with a spreadsheet (streamed)
POST	`/spreadsheet-search/store/`	Store a spreadsheet-chat reply

POST `/semantic-search/upload/`

Upload a document

Stores the file and, unless train_after_upload is false, queues it for indexing (text extraction, chunking, embedding). The response is immediate; follow progress with documents/status/ or the document's processing_state.

Supported types include PDF, Word, PowerPoint, text/Markdown, images (with ocr) and CSV/Excel. Files are virus-scanned; a rejected file returns 400 with reasons. Storage quotas apply when the deployment enforces them.

Re-uploading a SharePoint file (sharepoint_drive_item_id set) that already exists simply shares the existing document with you and returns 200.

Auth: Session token · In the app: My Data → Upload

Request body (multipart/form-data)

Field	Type	Required	Description
`path`	string (binary)	yes	The file.
`folder`	string	no	Virtual folder path, e.g. `Finance/2026`. Default root.
`train_after_upload`	boolean	no	`false` stores the file without indexing (state `untrained`). Default: `true`.
`ocr`	boolean	no	Run OCR on scanned PDFs and images. Default: `false`.
`sharepoint_drive_item_id`	string	no	Set by the SharePoint importer.

text

path: @Q3-report.pdf
folder: Finance/2026
ocr: false

Response 200 — SharePoint file already imported — shared with you instead.

Field	Type	Description
`uid`	string (uuid)
`folder`	string	Virtual folder path (`""` = root). Hidden on documents shared to you.
`path`	string	Media URL of the file (`/media/<org>/semantic-search/<user>/<name>`). Sign it to download.
`page_number`	integer	Pages processed so far (also total pages once complete).
`labels`	integer
`doc_class`	string	Document class inferred when the summary is generated (e.g. `contract`).
`status`	string	Human-readable processing status text.
`processing_state`	string (`"stored"`, `"pending_publish"`, `"queued"`, `"processing"`, `"completed"`, `"failed"`, `"publish_failed"`, `"untrained"`)	Machine state. `completed` means the file is indexed and chat-ready; `untrained` means it was uploaded with training off.
`enqueue_attempt_count`	integer
`last_error`	string
`queued_at`	string (date-time), nullable
`processing_started_at`	string (date-time), nullable
`completed_at`	string (date-time), nullable
`from_dms`	boolean
`sharepoint_drive_item_id`	string, nullable	Set when the file was imported from SharePoint/OneDrive.
`created_at`	string (date-time)

json

{
  "uid": "1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b",
  "folder": "Finance/2026",
  "path": "/media/Acme/semantic-search/jane/Q3-report.pdf",
  "page_number": 24,
  "labels": 1,
  "doc_class": "report",
  "status": "100% - Complete",
  "processing_state": "completed",
  "enqueue_attempt_count": 1,
  "last_error": "",
  "queued_at": "2026-09-20T09:00:02Z",
  "processing_started_at": "2026-09-20T09:00:05Z",
  "completed_at": "2026-09-20T09:01:40Z",
  "from_dms": false,
  "sharepoint_drive_item_id": null,
  "created_at": "2026-09-20T09:00:01Z"
}

Response 201 — Stored and queued (or stored untrained).

Field	Type	Description
`uid`	string (uuid)
`folder`	string	Virtual folder path (`""` = root). Hidden on documents shared to you.
`path`	string	Media URL of the file (`/media/<org>/semantic-search/<user>/<name>`). Sign it to download.
`page_number`	integer	Pages processed so far (also total pages once complete).
`labels`	integer
`doc_class`	string	Document class inferred when the summary is generated (e.g. `contract`).
`status`	string	Human-readable processing status text.
`processing_state`	string (`"stored"`, `"pending_publish"`, `"queued"`, `"processing"`, `"completed"`, `"failed"`, `"publish_failed"`, `"untrained"`)	Machine state. `completed` means the file is indexed and chat-ready; `untrained` means it was uploaded with training off.
`enqueue_attempt_count`	integer
`last_error`	string
`queued_at`	string (date-time), nullable
`processing_started_at`	string (date-time), nullable
`completed_at`	string (date-time), nullable
`from_dms`	boolean
`sharepoint_drive_item_id`	string, nullable	Set when the file was imported from SharePoint/OneDrive.
`created_at`	string (date-time)
`upload_id`	string	Correlation id for the processing pipeline.
`queue_pending`	boolean	Present when the queue was unavailable; the file will be retried automatically.
`queue_error`	string

json

{
  "uid": "1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b",
  "folder": "Finance/2026",
  "path": "/media/Acme/semantic-search/jane/Q3-report.pdf",
  "page_number": 1,
  "labels": 1,
  "doc_class": "",
  "status": "Queued",
  "processing_state": "queued",
  "enqueue_attempt_count": 1,
  "last_error": "",
  "queued_at": "2026-09-22T07:10:02Z",
  "processing_started_at": null,
  "completed_at": null,
  "from_dms": false,
  "sharepoint_drive_item_id": null,
  "created_at": "2026-09-22T07:10:01Z",
  "upload_id": "8b1f2c3d"
}

Response 400 — Validation error or the file failed the security scan.

json

{
  "detail": "File failed security scan.",
  "reasons": [
    "Executable content detected"
  ]
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/upload/" \
  -H "Authorization: Token $FINBLADE_TOKEN" \
  -F "path=@Q3-report.pdf" \
  -F "folder=Finance/2026"

POST `/semantic-search/upload/batch/`

Upload several documents at once

Multiple files under the files field, one folder for all. Each file is scanned and queued independently; the response reports per-file outcomes and a batch_id to poll.

Auth: Session token

Request body (multipart/form-data)

Field	Type	Required	Description
`files`	string (binary)[]	yes
`folder`	string	no
`ocr`	boolean	no	Default: `false`.

Response 202 — Accepted.

Field	Type	Description
`batch_id`	string (uuid)
`total`	integer
`results`	object[]
`results[].filename`	string
`results[].uid`	string
`results[].upload_id`	string
`results[].status`	string (`"queued"`, `"pending_publish"`, `"untrained"`, `"blocked"`, `"rejected"`)
`results[].reasons`	string[]
`results[].errors`	object

json

{
  "batch_id": "6f1e2d3c-4b5a-4c6d-8e9f-0a1b2c3d4e5f",
  "total": 2,
  "results": [
    {
      "filename": "a.pdf",
      "uid": "1b6c…",
      "upload_id": "8b1f2c3d",
      "status": "queued"
    },
    {
      "filename": "b.exe",
      "status": "blocked",
      "reasons": [
        "Executable content detected"
      ]
    }
  ]
}

Response 400 — No files.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/upload/batch/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/semantic-search/upload/batch/{batch_id}/status/`

Status of a batch upload

Auth: Session token

Path parameters

Field	Type	Required	Description
`batch_id`	string (uuid)	yes

Response 200 — Batch progress.

Field	Type	Description
`batch_id`	string
`state`	string (`"accepted"`, `"processing"`, `"completed"`, `"partial"`, `"failed"`)
`source_type`	string (`"direct"`, `"sharepoint"`, `"sql"`)
`total`	integer
`completed`	integer
`failed`	integer
`pending`	integer
`created_at`	string (date-time)
`completed_at`	string (date-time), nullable
`items`	object[]
`items[].id`	string
`items[].filename`	string
`items[].state`	string
`items[].last_error`	string
`items[].queued_at`	string, nullable
`items[].completed_at`	string, nullable

Response 404 — Not your batch.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X GET "https://api.example.com/semantic-search/upload/batch/<batch_id>/status/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/semantic-search/documents/`

List my documents

Documents you own plus documents shared with you, newest first, excluding the trash. folder is omitted on shared documents.

Auth: Session token · In the app: My Data file table

Response 200 — Documents.

Array of:

Field	Type	Description
`uid`	string (uuid)
`folder`	string	Virtual folder path (`""` = root). Hidden on documents shared to you.
`path`	string	Media URL of the file (`/media/<org>/semantic-search/<user>/<name>`). Sign it to download.
`page_number`	integer	Pages processed so far (also total pages once complete).
`labels`	integer
`doc_class`	string	Document class inferred when the summary is generated (e.g. `contract`).
`status`	string	Human-readable processing status text.
`processing_state`	string (`"stored"`, `"pending_publish"`, `"queued"`, `"processing"`, `"completed"`, `"failed"`, `"publish_failed"`, `"untrained"`)	Machine state. `completed` means the file is indexed and chat-ready; `untrained` means it was uploaded with training off.
`enqueue_attempt_count`	integer
`last_error`	string
`queued_at`	string (date-time), nullable
`processing_started_at`	string (date-time), nullable
`completed_at`	string (date-time), nullable
`from_dms`	boolean
`sharepoint_drive_item_id`	string, nullable	Set when the file was imported from SharePoint/OneDrive.
`created_at`	string (date-time)

Example

bash

curl -X GET "https://api.example.com/semantic-search/documents/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/semantic-search/documents/status/`

Documents still processing

Every visible document in a non-terminal state (stored, pending_publish, queued, processing). Cheap; meant for polling.

Auth: Session token · Rate limit: 120/min (own scope) · In the app: My Data status badges (polled every few seconds while uploads run)

Response 200 — Minimal rows.

Array of:

Field	Type	Description
`uid`	string
`processing_state`	string
`status`	string
`last_error`	string
`created_at`	string (date-time)

json

[
  {
    "uid": "1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b",
    "processing_state": "processing",
    "status": "45% - Embedding",
    "last_error": "",
    "created_at": "2026-09-22T07:10:01Z"
  }
]

Example

bash

curl -X GET "https://api.example.com/semantic-search/documents/status/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/documents/status/`

Status of specific documents

Current state of up to 500 documents by uid, whatever state they are in.

Auth: Session token · Rate limit: 120/min (own scope)

Request body (application/json)

Field	Type	Required	Description
`uids`	string (uuid)[]	yes

json

{
  "uids": [
    "1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b"
  ]
}

Response 200 — Minimal rows (unknown uids are omitted).

Array of:

Field	Type	Description
`uid`	string
`processing_state`	string
`status`	string
`last_error`	string
`created_at`	string (date-time)

Response 400 — uids not a list, or more than 500.

json

{
  "error": "path is required"
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/documents/status/" \
  -H "Authorization: Token $FINBLADE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"uids":["1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b"]}'

GET `/semantic-search/documents/{uid}/`

Get one document

Auth: Session token

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Response 200 — The document (no folder if it is only shared with you).

Field	Type	Description
`uid`	string (uuid)
`folder`	string	Virtual folder path (`""` = root). Hidden on documents shared to you.
`path`	string	Media URL of the file (`/media/<org>/semantic-search/<user>/<name>`). Sign it to download.
`page_number`	integer	Pages processed so far (also total pages once complete).
`labels`	integer
`doc_class`	string	Document class inferred when the summary is generated (e.g. `contract`).
`status`	string	Human-readable processing status text.
`processing_state`	string (`"stored"`, `"pending_publish"`, `"queued"`, `"processing"`, `"completed"`, `"failed"`, `"publish_failed"`, `"untrained"`)	Machine state. `completed` means the file is indexed and chat-ready; `untrained` means it was uploaded with training off.
`enqueue_attempt_count`	integer
`last_error`	string
`queued_at`	string (date-time), nullable
`processing_started_at`	string (date-time), nullable
`completed_at`	string (date-time), nullable
`from_dms`	boolean
`sharepoint_drive_item_id`	string, nullable	Set when the file was imported from SharePoint/OneDrive.
`created_at`	string (date-time)

json

{
  "uid": "1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b",
  "folder": "Finance/2026",
  "path": "/media/Acme/semantic-search/jane/Q3-report.pdf",
  "page_number": 24,
  "labels": 1,
  "doc_class": "report",
  "status": "100% - Complete",
  "processing_state": "completed",
  "enqueue_attempt_count": 1,
  "last_error": "",
  "queued_at": "2026-09-20T09:00:02Z",
  "processing_started_at": "2026-09-20T09:00:05Z",
  "completed_at": "2026-09-20T09:01:40Z",
  "from_dms": false,
  "sharepoint_drive_item_id": null,
  "created_at": "2026-09-20T09:00:01Z"
}

Response 404 — Not found or no access.

json

{
  "error": "path is required"
}

Example

bash

curl -X GET "https://api.example.com/semantic-search/documents/1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

PUT `/semantic-search/{uid}/update/`

Move a document to a folder

Only folder can be changed (renaming files is not supported). Owner only.

Auth: Session token · In the app: My Data → drag & drop / Move

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Request body (application/json)

Field	Type	Required	Description
`folder`	string	yes	New folder path (`""` for root).

json

{
  "folder": "Finance/2025"
}

Response 200 — Updated document.

Field	Type	Description
`uid`	string (uuid)
`folder`	string	Virtual folder path (`""` = root). Hidden on documents shared to you.
`path`	string	Media URL of the file (`/media/<org>/semantic-search/<user>/<name>`). Sign it to download.
`page_number`	integer	Pages processed so far (also total pages once complete).
`labels`	integer
`doc_class`	string	Document class inferred when the summary is generated (e.g. `contract`).
`status`	string	Human-readable processing status text.
`processing_state`	string (`"stored"`, `"pending_publish"`, `"queued"`, `"processing"`, `"completed"`, `"failed"`, `"publish_failed"`, `"untrained"`)	Machine state. `completed` means the file is indexed and chat-ready; `untrained` means it was uploaded with training off.
`enqueue_attempt_count`	integer
`last_error`	string
`queued_at`	string (date-time), nullable
`processing_started_at`	string (date-time), nullable
`completed_at`	string (date-time), nullable
`from_dms`	boolean
`sharepoint_drive_item_id`	string, nullable	Set when the file was imported from SharePoint/OneDrive.
`created_at`	string (date-time)

json

{
  "uid": "1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b",
  "folder": "Finance/2026",
  "path": "/media/Acme/semantic-search/jane/Q3-report.pdf",
  "page_number": 24,
  "labels": 1,
  "doc_class": "report",
  "status": "100% - Complete",
  "processing_state": "completed",
  "enqueue_attempt_count": 1,
  "last_error": "",
  "queued_at": "2026-09-20T09:00:02Z",
  "processing_started_at": "2026-09-20T09:00:05Z",
  "completed_at": "2026-09-20T09:01:40Z",
  "from_dms": false,
  "sharepoint_drive_item_id": null,
  "created_at": "2026-09-20T09:00:01Z"
}

Response 400 — A rename was attempted.

json

{
  "error": "Renaming is not supported from this endpoint."
}

Response 404 — Not your document.

json

{
  "error": "path is required"
}

Example

bash

curl -X PUT "https://api.example.com/semantic-search/<uid>/update/" \
  -H "Authorization: Token $FINBLADE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"folder":"Finance/2025"}'

DELETE `/semantic-search/{uid}/delete/`

Move a document to the trash

Soft delete. The file stays recoverable for 30 days (SEMANTIC_TRASH_RETENTION_DAYS) and is then purged with its index. Owner only.

Auth: Session token · In the app: My Data → Delete

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Response 200 — Trashed.

Field	Type	Description
`status`	string (`"trashed"`)
`uid`	string
`deleted_at`	string (date-time)
`purge_at`	string (date-time)

json

{
  "status": "trashed",
  "uid": "1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b",
  "deleted_at": "2026-09-22T07:30:00Z",
  "purge_at": "2026-10-22T07:30:00Z"
}

Response 404 — Not your document (or already trashed).

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X DELETE "https://api.example.com/semantic-search/1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b/delete/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/semantic-search/documents/trash/`

List trashed documents

Auth: Session token

Response 200 — Trashed documents with deleted_at and purge_at.

Example

bash

curl -X GET "https://api.example.com/semantic-search/documents/trash/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/{uid}/restore/`

Restore a document from the trash

Restores the row and re-queues the file for indexing.

Auth: Session token

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Response 200 — Restored.

Field	Type	Description
`status`	string (`"restored"`)
`uid`	string
`detail`	string

Response 404 — No such trashed document.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/<uid>/restore/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

DELETE `/semantic-search/{uid}/permanent-delete/`

Delete a document permanently

Removes the file, its index and its row immediately. Cannot be undone.

Auth: Session token

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Response 200 — Purged.

Field	Type	Description
`status`	string (`"permanently_deleted"`)
`uid`	string

Response 404 — Not your document.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X DELETE "https://api.example.com/semantic-search/<uid>/permanent-delete/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

DELETE `/semantic-search/folder/delete/`

Move a folder to the trash

Trashes every owned document in the folder and its sub-folders. folder may be sent in the body or as a query parameter; "" targets files at the root.

Auth: Session token · In the app: My Data → folder → Delete

Query parameters

Field	Type	Required	Description
`folder`	string	no

Response 200 — Trashed.

Field	Type	Description
`status`	string (`"trashed"`)
`count`	integer

json

{
  "status": "trashed",
  "count": 7
}

Response 400 — folder missing.

json

{
  "error": "path is required"
}

Response 404 — Folder empty or unknown.

json

{
  "error": "path is required"
}

Example

bash

curl -X DELETE "https://api.example.com/semantic-search/folder/delete/?folder=Finance%2F2025" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/folder/move/`

Move or rename a folder

Atomically rewrites the folder path of every owned document under from to sit under to. A rename is a move that keeps the same parent.

Auth: Session token · In the app: My Data → folder → Rename / Move

Request body (application/json)

Field	Type	Required	Description
`from`	string	yes
`to`	string	yes

json

{
  "from": "Finance/2025",
  "to": "Archive/Finance-2025"
}

Response 200 — Moved.

Field	Type	Description
`status`	string (`"moved"`)
`count`	integer
`from`	string
`to`	string

Response 400 — Missing from, or moving a folder into itself.

json

{
  "error": "path is required"
}

Response 404 — Folder empty or unknown.

json

{
  "error": "path is required"
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/folder/move/" \
  -H "Authorization: Token $FINBLADE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"from":"Finance/2025","to":"Archive/Finance-2025"}'

GET `/semantic-search/documents/{uid}/users/`

Who a document is shared with

Auth: Session token · In the app: Share dialog

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Response 200 — Owner and users.

Field	Type	Description
`owner`	object
`owner.id`	integer
`owner.username`	string
`owner.email`	string
`users`	object[]
`users[].id`	integer
`users[].username`	string
`users[].email`	string

json

{
  "owner": {
    "id": 42,
    "username": "jane",
    "email": "jane@acme.com"
  },
  "users": [
    {
      "id": 57,
      "username": "omar",
      "email": "omar@acme.com"
    }
  ]
}

Response 404 — Unknown document.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X GET "https://api.example.com/semantic-search/documents/<uid>/users/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/documents/{uid}/users/add/`

Share a document with users

Adds users by email. For SharePoint-sourced documents the sharing is mirrored to SharePoint in the background.

Auth: Session token · In the app: Share dialog → Add

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Request body (application/json)

Field	Type	Required	Description
`emails`	string (email)[]	yes

json

{
  "emails": [
    "omar@acme.com",
    "sara@acme.com"
  ]
}

Response 200 — Result.

Field	Type	Description
`added`	string[]
`not_found`	string[]
`detail`	string

json

{
  "added": [
    "omar@acme.com"
  ],
  "not_found": [
    "sara@acme.com"
  ],
  "detail": "1 user(s) added. 1 email(s) not found."
}

Response 400 — No emails.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/documents/<uid>/users/add/" \
  -H "Authorization: Token $FINBLADE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"emails":["omar@acme.com","sara@acme.com"]}'

POST `/semantic-search/documents/{uid}/users/remove/`

Stop sharing a document with users

Auth: Session token · In the app: Share dialog → Remove

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Request body (application/json)

Field	Type	Required	Description
`emails`	string (email)[]	yes

Response 200 — Result.

Field	Type	Description
`removed`	string[]
`not_found`	string[]
`detail`	string

Example

bash

curl -X POST "https://api.example.com/semantic-search/documents/<uid>/users/remove/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/semantic-search/documents/requests/`

Access requests

Without my, requests for your documents (to approve or deny). With my=1, requests you made.

Auth: Session token · In the app: Notifications → Access requests

Query parameters

Field	Type	Required	Description
`my`	string	no	Any value switches to requests you made.

Response 200 — Requests.

Array of:

Field	Type	Description
`id`	integer
`document`	object	A file in the semantic-search store.
`document.uid`	string (uuid)
`document.folder`	string	Virtual folder path (`""` = root). Hidden on documents shared to you.
`document.path`	string	Media URL of the file (`/media/<org>/semantic-search/<user>/<name>`). Sign it to download.
`document.page_number`	integer	Pages processed so far (also total pages once complete).
`document.labels`	integer
`document.doc_class`	string	Document class inferred when the summary is generated (e.g. `contract`).
`document.status`	string	Human-readable processing status text.
`document.processing_state`	string (`"stored"`, `"pending_publish"`, `"queued"`, `"processing"`, `"completed"`, `"failed"`, `"publish_failed"`, `"untrained"`)	Machine state. `completed` means the file is indexed and chat-ready; `untrained` means it was uploaded with training off.
`document.enqueue_attempt_count`	integer
`document.last_error`	string
`document.queued_at`	string (date-time), nullable
`document.processing_started_at`	string (date-time), nullable
`document.completed_at`	string (date-time), nullable
`document.from_dms`	boolean
`document.sharepoint_drive_item_id`	string, nullable	Set when the file was imported from SharePoint/OneDrive.
`document.created_at`	string (date-time)
`user`	object	The requester.
`user.id`	integer
`user.first_name`	string
`user.last_name`	string
`user.username`	string
`user.email`	string
`user.phone`	integer, nullable
`status`	string (`"pending"`, `"approved"`, `"denied"`, `"revoked"`)
`created_at`	string (date-time)
`updated_at`	string (date-time)

Example

bash

curl -X GET "https://api.example.com/semantic-search/documents/requests/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/documents/request/`

Request access to a document

Identify the document by its media path. Duplicate pending/approved requests are rejected.

Auth: Session token · In the app: Shared files → Request access

Request body (application/json)

Field	Type	Required	Description
`path`	string	yes	The document's media path (`/media/...` or `media/...`).

json

{
  "path": "/media/Acme/semantic-search/omar/policy.pdf"
}

Response 201 — Requested.

Field	Type	Description
`detail`	string	Human-readable explanation.
`code`	string	Machine-readable error code (present on some responses).

json

{
  "detail": "Document access requested"
}

Response 400 — A request already exists.

json

{
  "detail": "Authentication credentials were not provided."
}

Response 404 — No document at that path.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/documents/request/" \
  -H "Authorization: Token $FINBLADE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"path":"/media/Acme/semantic-search/omar/policy.pdf"}'

DELETE `/semantic-search/documents/request/`

Cancel my pending request

Auth: Session token

Request body (application/json)

Field	Type	Required	Description
`request_id`	integer	yes

Response 200 — Deleted (or, when not pending, {"error": "Document status is not pending"}).

Field	Type	Description
`status`	string

json

{
  "status": "deleted"
}

Example

bash

curl -X DELETE "https://api.example.com/semantic-search/documents/request/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/documents/approve/`

Approve an access request

Owner only. Shares the document with the requester and notifies them.

Auth: Session token · In the app: Notifications → Access requests → Approve

Request body (application/json)

Field	Type	Required	Description
`request_id`	integer	yes

Response 200 — Approved.

Field	Type	Description
`detail`	string	Human-readable explanation.
`code`	string	Machine-readable error code (present on some responses).

json

{
  "detail": "Document request approved"
}

Response 400 — Already processed.

json

{
  "detail": "Authentication credentials were not provided."
}

Response 404 — Not found or you are not the owner.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/documents/approve/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/documents/deny/`

Deny an access request

Auth: Session token · In the app: Notifications → Access requests → Deny

Request body (application/json)

Field	Type	Required	Description
`request_id`	integer	yes

Response 200 — Denied.

Field	Type	Description
`detail`	string	Human-readable explanation.
`code`	string	Machine-readable error code (present on some responses).

json

{
  "detail": "Document request denied"
}

Response 400 — Already processed.

json

{
  "detail": "Authentication credentials were not provided."
}

Response 404 — Not found or you are not the owner.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/documents/deny/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/documents/revoke/`

Revoke previously granted access

Auth: Session token · In the app: Notifications → Access requests → Revoke

Request body (application/json)

Field	Type	Required	Description
`request_id`	integer	yes

Response 200 — Revoked.

Field	Type	Description
`detail`	string	Human-readable explanation.
`code`	string	Machine-readable error code (present on some responses).

json

{
  "detail": "User access revoked"
}

Response 404 — Not found or you are not the owner.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/documents/revoke/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/semantic-search/{uid}/summary/`

Document summary & suggested prompts

Auth: Session token · In the app: Document viewer → Summary

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Query parameters

Field	Type	Required	Description
`language`	string (`"en"`, `"ar"`)	no	Default: `"en"`.

Response 200 — Summary payload (available: false when none has been generated yet).

Field	Type	Description
`language`	string (`"en"`, `"ar"`)
`available`	boolean	Whether a summary exists for this language.
`summary`	string, nullable
`doc_class`	string
`questions`	string[]	Suggested prompts for the document.
`generated_at`	string (date-time), nullable
`template_summaries`	string[]	Summaries produced from prompt templates (`rewrite-summary/`).

json

{
  "language": "en",
  "available": true,
  "summary": "The Q3 report covers revenue growth of 12.4%, margin expansion …",
  "doc_class": "report",
  "questions": [
    "What drove the margin expansion?",
    "Which region underperformed?"
  ],
  "generated_at": "2026-09-20T09:02:10Z",
  "template_summaries": []
}

Response 404 — Not found or no access.

json

{
  "detail": "Document does not exist"
}

Example

bash

curl -X GET "https://api.example.com/semantic-search/<uid>/summary/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/{uid}/summary/generate/`

Generate the summary & prompts

Produces the summary, document class and suggested prompts for one language from the indexed text. Returns 200 exists if one is already there, 409 while another generation is running.

Auth: Session token · In the app: Document viewer → Generate summary

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Query parameters

Field	Type	Required	Description
`language`	string (`"en"`, `"ar"`)	no	Default: `"en"`.

Response 200 — Already existed.

Field	Type	Description
`status`	string (`"exists"`)
`language`	string (`"en"`, `"ar"`)
`available`	boolean	Whether a summary exists for this language.
`summary`	string, nullable
`doc_class`	string
`questions`	string[]	Suggested prompts for the document.
`generated_at`	string (date-time), nullable
`template_summaries`	string[]	Summaries produced from prompt templates (`rewrite-summary/`).

Response 201 — Generated.

Field	Type	Description
`status`	string (`"generated"`)
`language`	string (`"en"`, `"ar"`)
`available`	boolean	Whether a summary exists for this language.
`summary`	string, nullable
`doc_class`	string
`questions`	string[]	Suggested prompts for the document.
`generated_at`	string (date-time), nullable
`template_summaries`	string[]	Summaries produced from prompt templates (`rewrite-summary/`).

Response 400 — The document has no indexed content yet.

json

{
  "detail": "Authentication credentials were not provided."
}

Response 409 — Generation already in progress.

json

{
  "detail": "Authentication credentials were not provided."
}

Response 502 — Summary service unavailable.

json

{
  "detail": "Authentication credentials were not provided."
}

Response 503 — Chunk store temporarily unavailable — retry.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/<uid>/summary/generate/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/semantic-search/{uid}/questions/`

Suggested prompts for a document

Auth: Session token

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Query parameters

Field	Type	Required	Description
`language`	string (`"en"`, `"ar"`)	no	Default: `"en"`.

Response 200 — Prompts.

Array of:

Field	Type	Description
`content`	string
`created_at`	string (date-time)

Example

bash

curl -X GET "https://api.example.com/semantic-search/<uid>/questions/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/questions/`

Add suggested prompts to a document

Auth: Session token

Request body (application/json)

Field	Type	Required	Description
`uid`	string (uuid)	yes
`questions`	string[]	yes
`language`	string (`"en"`, `"ar"`)	no	Default: `"en"`.

json

{
  "uid": "1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b",
  "questions": [
    "What are the payment terms?"
  ]
}

Response 201 — Created.

Field	Type	Description
`status`	string
`count`	integer

Response 400 — Missing uid or questions.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/questions/" \
  -H "Authorization: Token $FINBLADE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"uid":"1b6c1c1e-9d0a-4b6e-9f2a-3c4d5e6f7a8b","questions":["What are the payment terms?"]}'

POST `/semantic-search/summaries/`

Template-generated summaries

Summaries produced with a prompt template for one document, or for one chat session.

Auth: Session token

Request body (application/json)

Field	Type	Required
`template_id`	integer	yes
`document_id`	string (uuid)	no
`session_id`	integer	no

Response 200 — Summaries.

Field	Type	Description
`summaries`	object[]
`summaries[].id`	integer
`summaries[].document`	string, nullable
`summaries[].template`	integer
`summaries[].session`	integer, nullable
`summaries[].content`	string
`summaries[].created_at`	string (date-time)

Response 400 — Neither document nor session given.

json

{
  "error": "path is required"
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/summaries/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/semantic-search/rewrite-summary/`

Generate a summary from a prompt template

Runs the template's prompt over a document (by file name), or over free text query bound to a document or a session, and stores the result.

Auth: Session token · In the app: My Data → Templates → Apply

Request body (application/json)

Field	Type	Required	Description
`template_id`	integer	yes
`document_name`	string	no	File name of one of your semantic-search documents.
`session_id`	integer	no
`query`	string	no	Text to summarise instead of the document (required with `session_id`).

json

{
  "template_id": 12,
  "document_name": "Q3-report.pdf"
}

Response 200 — The summary.

Field	Type	Description
`summary`	string

Response 400 — Missing document/session.

json

{
  "error": "path is required"
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/rewrite-summary/" \
  -H "Authorization: Token $FINBLADE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"template_id":12,"document_name":"Q3-report.pdf"}'

POST `/semantic-search/search/`

Semantic search across my documents

Vector search over the user's index; returns the matching chunks with scores. 204 when nothing scores above the threshold.

Auth: Session token · In the app: My Data → search bar (relevance slider)

Request body (application/json)

Field	Type	Required	Description
`query`	string	yes
`score_threshold`	number	yes	Minimum similarity (0–1).

json

{
  "query": "termination clause notice period",
  "score_threshold": 0.5
}

Response 200 — Results grouped by document (shape defined by the search service).

Field	Type	Description
`results`	object

Response 204 — No results.

Example

bash

curl -X POST "https://api.example.com/semantic-search/search/" \
  -H "Authorization: Token $FINBLADE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"query":"termination clause notice period","score_threshold":0.5}'

POST `/semantic-search/re-embed/`

Re-index documents

Drops the existing index of the given documents (or all of yours) and queues them again — use after changing the embedding model or when a document looks stale.

Auth: Session token

Request body (application/json, optional)

Field	Type	Required	Description
`uids`	string (uuid)[]	no	Omit to re-index everything you own.

Response 200 — All queued.

Field	Type	Description
`detail`	string
`queued_documents`	string[]
`failed_documents`	object[]

Response 207 — Some queued, some failed.

Response 404 — No documents.

json

{
  "detail": "Authentication credentials were not provided."
}

Response 503 — Nothing could be queued.

json

{
  "detail": "Authentication credentials were not provided."
}

Example

bash

curl -X POST "https://api.example.com/semantic-search/re-embed/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/database-search/upload/`

Upload a CSV/Excel for the database chat

Stores the spreadsheet and queues it for tabular indexing. Answer questions over it with POST /database-search/chat/.

Auth: Session token · In the app: My Data → Upload (spreadsheets)

Request body (multipart/form-data)

Field	Type	Required	Description
`path`	string (binary)	yes
`folder`	string	no

Response 201 — Stored and queued.

Field	Type	Description
`uid`	string (uuid)
`folder`	string
`path`	string	`/media/<org>/database-search/<user>/<name>`
`created_at`	string (date-time)
`upload_id`	string

Response 400 — Validation error.

json

{
  "username": [
    "A user with that username already exists."
  ]
}

Response 503 — Stored but could not be queued; retry later.

json

{
  "error": "path is required"
}

Example

bash

curl -X POST "https://api.example.com/database-search/upload/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/database-search/documents/`

List my database-search files

Auth: Session token

Response 200 — Files.

Array of:

Field	Type	Description
`uid`	string (uuid)
`folder`	string
`path`	string	`/media/<org>/database-search/<user>/<name>`
`created_at`	string (date-time)

Example

bash

curl -X GET "https://api.example.com/database-search/documents/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

PUT `/database-search/{uid}/update/`

Move a database-search file to a folder

Auth: Session token

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Request body (application/json)

Field	Type	Required	Description
`folder`	string	yes

Response 200 — Updated.

Field	Type	Description
`uid`	string (uuid)
`folder`	string
`path`	string	`/media/<org>/database-search/<user>/<name>`
`created_at`	string (date-time)

json

{
  "uid": "2c7d2d2f-0e1b-4c7f-8a3b-4d5e6f7a8b9c",
  "folder": "",
  "path": "/media/Acme/database-search/jane/sales-2026.csv",
  "created_at": "2026-09-18T14:12:00Z"
}

Response 404 — Not your file.

json

{
  "error": "path is required"
}

Example

bash

curl -X PUT "https://api.example.com/database-search/<uid>/update/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

DELETE `/database-search/{uid}/delete/`

Delete a database-search file

Immediate, permanent (no trash for this store).

Auth: Session token

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Response 200 — Deleted.

Field	Type	Description
`status`	string (`"deleted"`)

Example

bash

curl -X DELETE "https://api.example.com/database-search/<uid>/delete/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

DELETE `/database-search/folder/delete/`

Delete a database-search folder

Auth: Session token

Query parameters

Field	Type	Required	Description
`folder`	string	no

Response 200 — Deleted.

Field	Type	Description
`status`	string
`count`	integer
`queue_warnings`	string[]

Response 404 — Folder empty or unknown.

json

{
  "error": "path is required"
}

Example

bash

curl -X DELETE "https://api.example.com/database-search/folder/delete/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/database-search/folder/move/`

Move or rename a database-search folder

Auth: Session token

Request body (application/json)

Field	Type	Required	Description
`from`	string	yes
`to`	string	yes

Response 200 — Moved.

Field	Type	Description
`status`	string
`count`	integer
`from`	string
`to`	string

Example

bash

curl -X POST "https://api.example.com/database-search/folder/move/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/database-search/chat/`

Ask a question about a CSV/Excel file

Runs the question against one database-search file through the analysis service and returns the answer plus any generated charts/files. Unlike the document chats this call is not streamed; both the question and the answer are stored in the session automatically.

Auth: Session token · In the app: My Data → Chat with a spreadsheet

Request body (application/json)

Field	Type	Required	Description
`question`	string	yes
`doc_uid`	string (uuid)	yes
`filename`	string	yes	File name (used to locate the file on disk).
`session`	integer	no	Session id of type `database_search`.

json

{
  "question": "Total sales per region in Q2?",
  "doc_uid": "2c7d2d2f-0e1b-4c7f-8a3b-4d5e6f7a8b9c",
  "filename": "sales-2026.csv",
  "session": 9203
}

Response 200 — Answer.

Field	Type	Description
`result`	string	Answer text (Markdown).
`generated_files_urls`	string[]	Media URLs of charts/exports produced.

json

{
  "result": "Q2 sales by region: East 1.2M, West 0.9M, Central 0.7M.",
  "generated_files_urls": [
    "/media/Acme/database-search/jane/charts/q2-by-region.png"
  ]
}

Response 500 — Service not configured or returned an error (body carries details).

json

{
  "error": "path is required"
}

Response 503 — Analysis service unreachable.

json

{
  "error": "path is required"
}

Example

bash

curl -X POST "https://api.example.com/database-search/chat/" \
  -H "Authorization: Token $FINBLADE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"question":"Total sales per region in Q2?","doc_uid":"2c7d2d2f-0e1b-4c7f-8a3b-4d5e6f7a8b9c","filename":"sales-2026.csv","session":9203}'

POST `/database-search/store/`

Store a database-chat message

Manual storage for the database chat (the chat endpoint already stores both sides; use this for client-side edits or chart configs).

Auth: Session token

Request body (application/json)

Field	Type	Required	Description
`response`	string	no
`session`	integer	no
`user_message`	boolean	no
`images`	string[]	no
`csv_files`	string[]	no
`chart_config`	object	no	Persisted onto the session's `query`.
`doc_uid`	string	no

Response 200 — Stored.

Response 404 — Session not found.

json

{
  "error": "path is required"
}

Example

bash

curl -X POST "https://api.example.com/database-search/store/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/database-search/{uid}/thoughts/`

Analysis notes for a database-search file

Reasoning notes recorded while the file was indexed.

Auth: Session token

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Response 200 — Notes.

Array of:

Field	Type	Description
`id`	integer
`doc`	string
`content`	string
`created_at`	string (date-time)

Example

bash

curl -X GET "https://api.example.com/database-search/<uid>/thoughts/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/database-search/`

All my database-chat messages

Auth: Session token

Response 200 — Messages.

Array of:

Field	Type	Description
`id`	integer
`user`	integer
`session`	integer, nullable
`from_server`	boolean
`message`	string
`images`	string[]
`csv_files`	string[]
`created_at`	string (date-time)

Example

bash

curl -X GET "https://api.example.com/database-search/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/spreadsheet-search/upload/`

Upload a spreadsheet for the spreadsheet chat

Indexes the rows synchronously (row-wise embeddings). Legacy store used by the spreadsheet viewer.

Auth: Session token

Request body (multipart/form-data)

Field	Type	Required	Description
`path`	string (binary)	yes

Response 201 — Stored and indexed.

Field	Type	Description
`uid`	string
`path`	string
`created_at`	string (date-time)

Example

bash

curl -X POST "https://api.example.com/spreadsheet-search/upload/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

GET `/spreadsheet-search/documents/`

List my spreadsheet-search files

Auth: Session token

Response 200 — Files.

Array of:

Field	Type	Description
`uid`	string
`path`	string
`created_at`	string (date-time)

Example

bash

curl -X GET "https://api.example.com/spreadsheet-search/documents/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

DELETE `/spreadsheet-search/{uid}/delete/`

Delete a spreadsheet-search file

Auth: Session token

Path parameters

Field	Type	Required	Description
`uid`	string (uuid)	yes

Response 200 — Deleted.

Response 404 — Not found.

json

{
  "error": "path is required"
}

Example

bash

curl -X DELETE "https://api.example.com/spreadsheet-search/<uid>/delete/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/spreadsheet-search/chat/`

Chat with a spreadsheet (streamed)

Streams grouped sources followed by the answer, like the legacy semantic chat. Selects files by name.

Auth: Session token

Request body (application/json)

Field	Type	Required	Description
`question`	string	yes
`filename`	string	yes	Comma-separated file names.
`retriever`	string	no

Response 200 — Streamed text.

Content type: text/event-stream

Example

bash

curl -X POST "https://api.example.com/spreadsheet-search/chat/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

POST `/spreadsheet-search/store/`

Store a spreadsheet-chat reply

Auth: Session token

Request body (application/json)

Field	Type	Required	Description
`response`	string	yes
`sources`	string	yes	JSON-encoded sources array.

Response 200 — Stored.

Example

bash

curl -X POST "https://api.example.com/spreadsheet-search/store/" \
  -H "Authorization: Token $FINBLADE_TOKEN"

My Data — documents ​

Three stores, one shape ​

Processing status ​

Endpoints ​

POST /semantic-search/upload/ ​

POST /semantic-search/upload/batch/ ​

GET /semantic-search/upload/batch/{batch_id}/status/ ​

GET /semantic-search/documents/ ​

GET /semantic-search/documents/status/ ​

POST /semantic-search/documents/status/ ​

GET /semantic-search/documents/{uid}/ ​

PUT /semantic-search/{uid}/update/ ​

DELETE /semantic-search/{uid}/delete/ ​

GET /semantic-search/documents/trash/ ​

POST /semantic-search/{uid}/restore/ ​

DELETE /semantic-search/{uid}/permanent-delete/ ​

DELETE /semantic-search/folder/delete/ ​

POST /semantic-search/folder/move/ ​

GET /semantic-search/documents/{uid}/users/ ​

POST /semantic-search/documents/{uid}/users/add/ ​

POST /semantic-search/documents/{uid}/users/remove/ ​

GET /semantic-search/documents/requests/ ​

POST /semantic-search/documents/request/ ​

DELETE /semantic-search/documents/request/ ​

POST /semantic-search/documents/approve/ ​

POST /semantic-search/documents/deny/ ​

POST /semantic-search/documents/revoke/ ​

GET /semantic-search/{uid}/summary/ ​

POST /semantic-search/{uid}/summary/generate/ ​

GET /semantic-search/{uid}/questions/ ​

POST /semantic-search/questions/ ​

POST /semantic-search/summaries/ ​

POST /semantic-search/rewrite-summary/ ​

POST /semantic-search/search/ ​

POST /semantic-search/re-embed/ ​

POST /database-search/upload/ ​

GET /database-search/documents/ ​

PUT /database-search/{uid}/update/ ​

DELETE /database-search/{uid}/delete/ ​

DELETE /database-search/folder/delete/ ​

POST /database-search/folder/move/ ​

POST /database-search/chat/ ​

POST /database-search/store/ ​

GET /database-search/{uid}/thoughts/ ​

GET /database-search/ ​

POST /spreadsheet-search/upload/ ​

GET /spreadsheet-search/documents/ ​

DELETE /spreadsheet-search/{uid}/delete/ ​

POST /spreadsheet-search/chat/ ​

POST /spreadsheet-search/store/ ​

My Data — documents

Three stores, one shape

Processing status

Endpoints

POST `/semantic-search/upload/`

POST `/semantic-search/upload/batch/`

GET `/semantic-search/upload/batch/{batch_id}/status/`

GET `/semantic-search/documents/`

GET `/semantic-search/documents/status/`

POST `/semantic-search/documents/status/`

GET `/semantic-search/documents/{uid}/`

PUT `/semantic-search/{uid}/update/`

DELETE `/semantic-search/{uid}/delete/`

GET `/semantic-search/documents/trash/`

POST `/semantic-search/{uid}/restore/`

DELETE `/semantic-search/{uid}/permanent-delete/`

DELETE `/semantic-search/folder/delete/`

POST `/semantic-search/folder/move/`

GET `/semantic-search/documents/{uid}/users/`

POST `/semantic-search/documents/{uid}/users/add/`

POST `/semantic-search/documents/{uid}/users/remove/`

GET `/semantic-search/documents/requests/`

POST `/semantic-search/documents/request/`

DELETE `/semantic-search/documents/request/`

POST `/semantic-search/documents/approve/`

POST `/semantic-search/documents/deny/`

POST `/semantic-search/documents/revoke/`

GET `/semantic-search/{uid}/summary/`

POST `/semantic-search/{uid}/summary/generate/`

GET `/semantic-search/{uid}/questions/`

POST `/semantic-search/questions/`

POST `/semantic-search/summaries/`

POST `/semantic-search/rewrite-summary/`

POST `/semantic-search/search/`

POST `/semantic-search/re-embed/`

POST `/database-search/upload/`

GET `/database-search/documents/`

PUT `/database-search/{uid}/update/`

DELETE `/database-search/{uid}/delete/`

DELETE `/database-search/folder/delete/`

POST `/database-search/folder/move/`

POST `/database-search/chat/`

POST `/database-search/store/`

GET `/database-search/{uid}/thoughts/`

GET `/database-search/`

POST `/spreadsheet-search/upload/`

GET `/spreadsheet-search/documents/`

DELETE `/spreadsheet-search/{uid}/delete/`

POST `/spreadsheet-search/chat/`

POST `/spreadsheet-search/store/`