Local Open WebUI Environment Setup for Direct Windows OS Automation via Unrestricted AI Agent
Budget / Salary₹600–1,500
TypeFreelance project
LocationRemote
Posted1 hour ago
**Title:** Local Open WebUI Environment Setup for Direct Windows OS Automation via Unrestricted AI Agent
---
### **1. Technical Scope (To be executed via AnyDesk/Remote Desktop):**
1. **Native Host Environment (No Docker):**
* Install and configure **Open WebUI** directly on the **Windows Python native environment** (strictly avoiding Docker/containers).
* Ensure the application runs with full host-OS privileges to allow unrestricted access to Windows processes, desktop, and system command executions.
2. **Model Integration & Unrestricted Logic:**
* Connect Open WebUI to a local **`qwen2.5-coder`** model (via Ollama) or an equivalent uncensored/unrestricted coding model.
* Configure model parameters, Function Calling, and System Prompts to ensure high accuracy in understanding system commands and zero refusal behavior.
3. **Custom System Command Execution Tool:**
* Develop/integrate a **Custom Tool** within Open WebUI using Python/PowerShell (`subprocess`).
* Cleanly handle arguments and eliminate encoding/escaping bugs (such as HTML entities/`"`).
* Ensure the tool can launch browsers, open system applications, and manage files directly on the Windows host.
---
### **2. Deliverables & Acceptance Criteria (Strict Requirements):**
1. **Video Proof Demonstration (1-Minute Screen Recording):**
* Before finalizing the project, provide a short (approx. 1 minute) video recording showing the active Windows desktop.
* Demonstrate live command execution by asking the model in the chat (in English or Arabic) to trigger direct OS actions (e.g., *"Open YouTube in the browser"* or *"Launch Notepad"*).
* The video must clearly show Windows responding immediately by opening the target application/website on screen.
2. **Live Remote Verification:**
* Perform a live test during the remote session to verify seamless execution before completing the job.
---
### **1. Technical Scope (To be executed via AnyDesk/Remote Desktop):**
1. **Native Host Environment (No Docker):**
* Install and configure **Open WebUI** directly on the **Windows Python native environment** (strictly avoiding Docker/containers).
* Ensure the application runs with full host-OS privileges to allow unrestricted access to Windows processes, desktop, and system command executions.
2. **Model Integration & Unrestricted Logic:**
* Connect Open WebUI to a local **`qwen2.5-coder`** model (via Ollama) or an equivalent uncensored/unrestricted coding model.
* Configure model parameters, Function Calling, and System Prompts to ensure high accuracy in understanding system commands and zero refusal behavior.
3. **Custom System Command Execution Tool:**
* Develop/integrate a **Custom Tool** within Open WebUI using Python/PowerShell (`subprocess`).
* Cleanly handle arguments and eliminate encoding/escaping bugs (such as HTML entities/`"`).
* Ensure the tool can launch browsers, open system applications, and manage files directly on the Windows host.
---
### **2. Deliverables & Acceptance Criteria (Strict Requirements):**
1. **Video Proof Demonstration (1-Minute Screen Recording):**
* Before finalizing the project, provide a short (approx. 1 minute) video recording showing the active Windows desktop.
* Demonstrate live command execution by asking the model in the chat (in English or Arabic) to trigger direct OS actions (e.g., *"Open YouTube in the browser"* or *"Launch Notepad"*).
* The video must clearly show Windows responding immediately by opening the target application/website on screen.
2. **Live Remote Verification:**
* Perform a live test during the remote session to verify seamless execution before completing the job.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.