1. Application Overview

AdminT00Lz is an internal systems-administration toolkit built for the IT operations team. It consolidates day-to-day IT administration tasks into a single web-based portal, including:

• Active Directory user and group management
• Exchange Server administration (on-premises & online)
• Post-patch scanning and reporting of Windows servers
• Automated scan scheduling via background services
• Asset inventory tracking (servers, desktops, printers)
• Admin skills knowledge base
• API integration with external ticketing and task systems
• API key management and an interactive API request simulator

The application runs as a Windows Service and is accessible through a web browser, providing a familiar admin dashboard experience with role-based access control.

2. Architecture

Technology Stack

Authentication Model

The application uses dual authentication:

• Windows Authentication — HTTP.sys Negotiate (Kerberos/NTLM) for interactive browser sessions.
• JWT Bearer — Token-based authentication for API consumers, issued via the OAuth token endpoint.

Key Services

Network Configuration

HTTP and HTTPS ports are configurable through the application settings stored in SQLite. Default ports are 88 (HTTP) and 8443 (HTTPS).

3. Authentication & Access Control

Page-level access is enforced by the AccessControlFilter. Users are assigned one of three roles through access groups:

Users who are members of the configured LDAP admin group are treated as superusers and bypass all role checks.

Page Categories

• Tasks pages (AD Admin, Exchange Admin, Skills) — Require any assigned role (Administrator, Editor, or Viewer).
• Tools pages (API Request Simulator, Agent Dialog) — Require any assigned role (Administrator, Editor, or Viewer).
• Controls pages (Settings, Auth Config, API Key Management) — Require Administrator role.
• Logs pages (Task Execution Log) — Require Administrator role.

4. Features

4g. Agentic AI Features

adminT00Lz includes AI-powered capabilities that use Large Language Models (LLMs) to interpret natural language administrative requests and convert them into structured API calls. This enables administrators to interact with the system using plain language instead of learning specific API syntax.

Overview

The Agentic AI system consists of three main components:

• Agent Dialog — A conversational interface where administrators can describe tasks in natural language
• Task Skills — System prompts that teach the AI how to interpret requests and map them to API calls
• LLM Provider — The underlying AI engine (currently Ollama, with Azure OpenAI and OpenAI coming soon)

How It Works

1. Administrator describes a task in plain language in the Agent Dialog
2. The request is sent to the configured LLM Provider along with relevant Task Skills system prompts
3. The AI analyzes the request and generates a structured JSON payload with the appropriate task code and parameters
4. Administrator reviews the generated request and can send it directly to the Task API for execution

Example Workflow

{
  "TaskCode": "ADD_DL_MEMBER",
  "Parameters": {
    "DistributionList": "Marketing Team",
    "UserName": "john.doe"
  }
}

Requirements

• Administrator access — Only administrators can use AI features
• Configured LLM provider — Ollama must be installed and running, or Azure OpenAI credentials configured
• Defined Task Skills — At least one task skill must be configured with appropriate system prompts

Note: The AI assistant is a tool to help generate requests faster. Always review the generated JSON before sending it to the API to ensure it accurately represents your intent.

5. Controls & Settings

The Controls menu is available in the top navigation bar (not the sidebar) and is restricted to Administrator users. It contains application-wide configuration pages.

Authentication

Application

Agentic AI

Logs

6. API Reference

adminT00Lz exposes a REST API for programmatic integration with external systems.

Authentication Flow

1. Obtain an API key from the API Key Management page.
2. Request a JWT token from the OAuth endpoint:

   POST /api/oauth/token
   Content-Type: application/x-www-form-urlencoded
   
   grant_type=client_credentials&client_id=YOUR_KEY&client_secret=YOUR_SECRET

3. Include the token in subsequent API requests:

   Authorization: Bearer <token>

Task API

Use the API Request Simulator to test API calls interactively before integrating them into scripts or external applications.