From a4dd5f87c0c74cda79554d6a551e750fca0114df Mon Sep 17 00:00:00 2001
From: Rohit Rajan <rohit.rajan031101@gmail.com>
Date: Fri, 24 Apr 2026 13:36:07 +0530
Subject: [PATCH 1/3] feat: add support for document extraction

---
 client.py | 38 +++++++++++++++++++++++++++++++++++++-
 1 file changed, 37 insertions(+), 1 deletion(-)

diff --git a/client.py b/client.py
index 3d371cb..2daa8d9 100644
--- a/client.py
+++ b/client.py
@@ -1,7 +1,8 @@
 import time
+import os
 import httpx
 from datetime import datetime, timezone
-from typing import Optional
+from typing import Optional, Union
 from .types import Config, MaxunError
 
 
@@ -161,6 +162,41 @@ async def extract_with_llm(self, options: dict):
             self.client.post("/extract/llm", json=options, timeout=300)
         )
 
+    async def create_document_robot(
+        self,
+        file: Union[str, bytes],
+        prompt: str,
+        robot_name: Optional[str] = None,
+        ollama_model: Optional[str] = None,
+        file_name: Optional[str] = None,
+    ) -> dict:
+        """Create a document-extraction robot from a PDF file path or bytes."""
+        if isinstance(file, str):
+            file_name = file_name or os.path.basename(file)
+            with open(file, 'rb') as f:
+                file_bytes = f.read()
+        else:
+            file_bytes = file
+            file_name = file_name or 'document.pdf'
+
+        data = {'prompt': prompt}
+        if robot_name:
+            data['robotName'] = robot_name
+        if ollama_model:
+            data['ollamaModel'] = ollama_model
+
+        response = await self.client.post(
+            '/robots/document',
+            files={'file': (file_name, file_bytes, 'application/pdf')},
+            data=data,
+            timeout=120,
+        )
+        response.raise_for_status()
+        body = response.json()
+        if not body.get('data') and not body.get('robot'):
+            raise MaxunError('Failed to create document robot')
+        return body
+
     async def create_crawl_robot(self, url: str, options: dict):
         return await self._handle(
             self.client.post("/crawl", json={"url": url, **options})

From bdafdcd87e637389e65b7921317e3fbdf1ec8b89 Mon Sep 17 00:00:00 2001
From: Rohit Rajan <rohit.rajan031101@gmail.com>
Date: Thu, 7 May 2026 17:22:56 +0530
Subject: [PATCH 2/3] feat: add document parsing support

---
 client.py | 44 +++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 43 insertions(+), 1 deletion(-)

diff --git a/client.py b/client.py
index 2daa8d9..8124e55 100644
--- a/client.py
+++ b/client.py
@@ -12,7 +12,8 @@ def __init__(self, config: Config):
 
         headers = {
             "x-api-key": self.api_key,
-            "Content-Type": "application/json",
+            # Content-Type is intentionally omitted here so httpx can set it
+            # correctly per request (e.g. multipart/form-data for file uploads)
         }
 
         if config.team_id:
@@ -197,6 +198,47 @@ async def create_document_robot(
             raise MaxunError('Failed to create document robot')
         return body
 
+    async def create_document_parse_robot(
+        self,
+        file: Union[str, bytes],
+        output_formats: list,
+        robot_name: Optional[str] = None,
+        file_name: Optional[str] = None,
+    ) -> dict:
+        """Create a document-parse robot from a PDF file path or bytes."""
+        if isinstance(file, str):
+            file_name = file_name or os.path.basename(file)
+            with open(file, 'rb') as f:
+                file_bytes = f.read()
+        else:
+            file_bytes = file
+            file_name = file_name or 'document.pdf'
+
+        valid_formats = {'markdown', 'html', 'links'}
+        filtered = [f for f in output_formats if f in valid_formats]
+        if not filtered:
+            raise MaxunError('At least one valid output format is required (markdown, html, links)')
+
+        data = {}
+        if robot_name:
+            data['robotName'] = robot_name
+
+        files_payload = [('file', (file_name, file_bytes, 'application/pdf'))]
+        for fmt in filtered:
+            files_payload.append(('outputFormats[]', (None, fmt)))
+
+        response = await self.client.post(
+            '/robots/document-parse',
+            files=files_payload,
+            data=data,
+            timeout=120,
+        )
+        response.raise_for_status()
+        body = response.json()
+        if not body.get('data') and not body.get('robot'):
+            raise MaxunError('Failed to create document-parse robot')
+        return body
+
     async def create_crawl_robot(self, url: str, options: dict):
         return await self._handle(
             self.client.post("/crawl", json={"url": url, **options})

From 30df5b4252b5ee787cd675268588c9e948c3a6bc Mon Sep 17 00:00:00 2001
From: Rohit Rajan <rohit.rajan031101@gmail.com>
Date: Fri, 8 May 2026 12:19:19 +0530
Subject: [PATCH 3/3] chore: rename doc methods

---
 client.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/client.py b/client.py
index 8124e55..9b686ac 100644
--- a/client.py
+++ b/client.py
@@ -163,7 +163,7 @@ async def extract_with_llm(self, options: dict):
             self.client.post("/extract/llm", json=options, timeout=300)
         )
 
-    async def create_document_robot(
+    async def create_document_extract_robot(
         self,
         file: Union[str, bytes],
         prompt: str,