Q-ACE: Agentic QA Framework

Modular AI Agent Engine for Quality Assurance tasks, built with FastAPI and modern LLM providers. Q-ACE transforms standard QA tools into intelligent agents capable of analysis, verification, and automated orchestration.

🚀 Key Features

Modular Architecture: Clean separation between core orchestration, handlers for tool logic, providers for LLMs, and tools for low-level execution.
Role-Based Access Control (RBAC): Secure login system with admin and tester roles, managing permissions for specific tools and analytics.
AI-Powered Agents:
- Browser Agent: High-level web automation with real-time feedback and AI analysis.
- Database Agent: Natural language query execution with AI verification of results.
- REST API Agent: Intelligent interaction with REST endpoints to fetch, create, edit, and verify data using natural language prompts.
- Test Generator Agent: Automated generation of comprehensive test cases (Functionality, Security, etc.) from specifications using advanced testing techniques.
Extended Tool Integration (Under Implementation):
- Built-in placeholders for Jenkins, GitHub Actions, Jira, Slack, Qase, Postman, and GitHub.
LLM Provider Switching: Support for Google Gemini, local Ollama, OpenAI, Anthropic, DeepSeek, and more.
Integrated Ollama Management: Manage your local Ollama server (Start/Stop/Status) directly from the interface.
Modern UI: A premium, responsive chat-based orchestration hub.

🛠 ️ Project Structure

core/: Orchestrator, Context Management, Tool Registry, Auth Utilities, and LLM Client.
handlers/: Logic layer for tools (SQLite, API, Spec Analyzer, Browser, and placeholders).
providers/: LLM provider implementations (Gemini, Ollama).
tools/: Low-level execution plugins.
static/: Modern HTML/JS frontend (Alpine.js, TailwindCSS).
data/: SQLite databases (auth.db), test specs, and Browser Agent documentation/history.
main.py: FastAPI application entry point.

🚦 Quick Start

1. Prerequisites

Python 3.9+
Ollama (for local LLM support)
Google API Key (for Gemini support)

2. Automatic Setup

Simply run the provided installer. It will handle virtual environments (including the specialized Browser Agent .venv), dependencies, and database initialization:

install.bat

3. Run the Server

Launch the framework after installation:

run.bat

Open http://localhost:8090 in your browser.

🌐 Specialized Documentation

Browser Agent Guide: Setup, configuration, and automation features.
Database Agent Guide: Text-to-SQL, sample databases, and verification.
REST API Agent Guide: Natural language API interaction and analysis.
Test Generator Agent Guide: Specification analysis and test case generation.

🔐 Authentication & Roles

The framework uses JWT-based authentication. Use init_auth_db.py to set up initial users:

Admin: Full access to all tools and administrative analytics.
Tester: Access limited to assigned tools (e.g., API, SQLite).

🧪 Extending the Framework

Adding a Handler

Handlers manage the high-level logic and UI definition for a tool. Create a new file in handlers/ inheriting from BaseHandler.

Adding a Tool

Tools provide the execution layer (e.g., making the actual HTTP request). Create a new file in tools/ inheriting from BaseTool.

from core.base_tool import BaseTool

class CustomTool(BaseTool):
    @property
    def name(self): return "custom_tool"

    async def execute(self, **kwargs):
        return {"status": "success", "data": "..."}

Built with ❤️ by ATID College

Docs