Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.spn.wtf/llms.txt

Use this file to discover all available pages before exploring further.

​
Order Drafts

Order drafts are WhatsApp-parsed orders pending review before conversion to real client orders. All routes are prefixed with /order-drafts.

​
GET /order-drafts

Get all pending order drafts. Response 
[
  {
    "id": 1,
    "created_at": "2025-01-15T10:30:00",
    "phone_number": "917595903437",
    "raw_message": "2 rings 5g each for Priya",
    "entries": [
      {
        "id": 1,
        "item_name": "Ring",
        "weight": 5.0,
        "quantity": 2,
        "client_name": "Priya"
      }
    ]
  }
]

​
GET /order-drafts/{draft_id}

Get a specific order draft with all its entries.

​
PUT /order-drafts/{draft_id}

Update metadata of an order draft.

​
DELETE /order-drafts/{draft_id}

Delete an order draft and all its entries.

​
PUT /order-drafts/{draft_id}/entries/{entry_id}

Update a specific entry within a draft. Request Body 
{
  "item_name": "Ring",
  "weight": 5.2,
  "quantity": 2,
  "client_id": 1,
  "slip_number": "SL-042"
}

​
POST /order-drafts/{draft_id}/recalculate_slip_numbers

Recalculate and reassign slip numbers for all entries in a draft.

​
POST /order-drafts/{draft_id}/convert

Convert a reviewed draft into actual client orders. Creates real client_order records and deletes the draft.

​
WhatsApp Webhook

​
POST /gupshup_incoming_webhook

Receives incoming WhatsApp messages from Gupshup. Flow:
  1. Customer sends an order via WhatsApp
  2. Gupshup forwards the message to this endpoint
  3. Sender’s phone number is checked against ALLOWED_PHONE_NUMBERS
  4. If allowed, message is sent to OpenAI for structured parsing
  5. An OrderDraft is created with parsed entries
  6. A WhatsApp confirmation reply is sent back
  7. Staff review the draft at /order-drafts in the frontend
  8. Once confirmed, draft is converted to real client orders
Only phone numbers listed in ALLOWED_PHONE_NUMBERS env var are processed. All others are silently ignored.

​
Khata Upload

Khata upload processes scanned/photographed ledger sheets using OCR.

​
POST /khata-upload

Upload a khata image/PDF and get back parsed entries (v1). Request: multipart/form-data
FieldTypeDescription
filefileImage or PDF of the khata page

​
POST /khata-upload-v2

Upload and process using the v2 parsing pipeline.

​
POST /khata-submit

Submit reviewed khata entries into the ledger system.

​
POST /khata/submit

Alternative submission endpoint (v2 flow).