NDIM Engine v0.10.0-alpha.1

NDIM Engine Alpha Testing Build

This is an unsigned alpha testing build for invited researchers and institutional testers.

NDIM Engine runs locally on your computer. Core functions do not require internet:

• narrative intake,
• CSV/PDF/text import,
• deterministic encoding,
• evidence ledger,
• modelling,
• policy export,
• stress-test corpus,
• backup and restore.

Only optional remote LLM augmentation requires internet and a user-provided key.

Why your computer may show a warning

This alpha build is not yet code-signed.

That means Windows or macOS may warn that the publisher cannot be verified. This is expected for an invited alpha build and does not mean the app is malicious.

Only download NDIM Engine from the official GitHub Releases page. Do not install builds sent through chat apps or email attachments unless the file hash matches the release checksum.

Windows: easiest alpha path

Preferred file:

NDIM-Engine-Windows-Portable-<version>.zip

Steps:

1. Download the portable ZIP from GitHub Releases.
2. Extract the ZIP to a normal folder, such as Documents.
3. Open the extracted folder.
4. Double-click Launch NDIM Engine.cmd.
5. If Windows shows a warning, choose More info and then Run anyway.
6. In the launcher, click Check offline readiness.

Optional file:

NDIM-Engine-Setup-<version>.exe

Use this only if you want a Start Menu shortcut.

macOS: alpha path

Preferred file:

NDIM-Engine-<version>.dmg

Steps:

1. Download the DMG from GitHub Releases.
2. Open it and drag NDIM Engine into Applications if shown.
3. If macOS blocks the app, right-click NDIM Engine and choose Open.
4. Confirm that you want to open it.
5. Use Check offline readiness after the launcher opens.

For broad public distribution, macOS builds should later be signed and notarized.

Linux: alpha path

Preferred files:

• ndim-engine_<version>_amd64.deb for Debian/Ubuntu systems
• AppImage when available

For .deb:

sudo apt install ./ndim-engine_<version>_amd64.deb
ndim-engine

For unpacked alpha artifacts, run the launcher from the extracted folder.

Verify checksums

Each alpha release should include:

SHA256SUMS.txt

This file contains fingerprints for release artifacts. It lets testers confirm that downloads were not changed after release.

Windows PowerShell:

Get-FileHash .\NDIM-Engine-Windows-Portable-<version>.zip -Algorithm SHA256

macOS/Linux:

shasum -a 256 NDIM-Engine-<version>.dmg
sha256sum ndim-engine_<version>_amd64.deb

Compare the result with SHA256SUMS.txt.

What to send back when something fails

Open the launcher and click:

Export support bundle

The support bundle excludes the narrative database by default. Full backups are sensitive and should only be shared intentionally.

Demo walkthrough

After launching the app, testers can open:

http://127.0.0.1:8010/demo-video

This is a browser-based walkthrough of the current NDIM workflow. It is not a separate app; it shows the intended testing path from workspace selection through narrative intake, encoding, modelling, digital twin feedback, inoculation, and policy export.

Alpha release boundary

This build is suitable for invited alpha testing.

It is not yet a fully signed public institutional release. Before broad deployment, NDIM Engine should add:

• Windows code signing,
• macOS signing and notarization,
• tested Linux release packages on target distributions.

NDIM Engine v0.9.0-alpha.1

NDIM Engine Desktop Release Notes

Package

This Windows package is a local-first NDIM Engine desktop distribution. It installs and runs the research tool on the user's own machine.

Files to publish

• dist/NDIM Engine Setup.exe - one-click Windows installer.
• dist/NDIM-Engine-Windows-Portable.zip - portable archive containing the bundled runtime. Users can unzip it and run Launch NDIM Engine.cmd.

What is bundled

• NDIM Engine backend and integrated workflow UI.
• Bundled Python runtime and installed Python packages required by the current app.
• Manual and stress-test instructions.
• Full stress_test_corpus/ CSV/TXT corpus.
• Local SQLite database support.

Users do not need to install Python, Node, npm, pip packages, or project requirements manually.

Installation

1. Download NDIM Engine Setup.exe.
2. Double-click the installer.
3. Choose an install folder or keep the default.
4. Launch NDIM Engine from the Start Menu shortcut.
5. The app opens a local browser page at http://127.0.0.1:8010 or the next available local port.

Portable option:

1. Download NDIM-Engine-Windows-Portable.zip.
2. Extract the ZIP.
3. Open NDIM Engine Runtime/Launch NDIM Engine.cmd.

Data storage

The app stores local research data on the user's machine. In desktop mode the launcher creates folders for data, uploads, exports, logs, SDMX packages, backups, and support bundles.

Stress-test corpus

The stress-test corpus is included in the package. It contains route-specific CSV files plus a bonus TXT narrative for testing narrative ingestion, SDMX readiness, approval, encoding, modelling, and policy-output workflow.

Distribution

GitHub Releases is the recommended place to host the installer. Do not rely on local file paths as download links; upload NDIM Engine Setup.exe and NDIM-Engine-Windows-Portable.zip as release assets.

Known limitations

• This Windows build is unsigned, so Windows SmartScreen may warn users on first launch.
• Advanced Torch/Pyro analytics are not bundled in this package; deterministic fallback analytics remain available.
• macOS and Linux packages must be built on those operating systems.

NDIM Engine v0.8.0-alpha.1

NDIM Engine Desktop Release Notes

Package

This Windows package is a local-first NDIM Engine desktop distribution. It installs and runs the research tool on the user's own machine.

Files to publish

• dist/NDIM Engine Setup.exe - one-click Windows installer.
• dist/NDIM-Engine-Windows-Portable.zip - portable archive containing the bundled runtime. Users can unzip it and run Launch NDIM Engine.cmd.

What is bundled

• NDIM Engine backend and integrated workflow UI.
• Bundled Python runtime and installed Python packages required by the current app.
• Manual and stress-test instructions.
• Full stress_test_corpus/ CSV/TXT corpus.
• Local SQLite database support.

Users do not need to install Python, Node, npm, pip packages, or project requirements manually.

Installation

1. Download NDIM Engine Setup.exe.
2. Double-click the installer.
3. Choose an install folder or keep the default.
4. Launch NDIM Engine from the Start Menu shortcut.
5. The app opens a local browser page at http://127.0.0.1:8010 or the next available local port.

Portable option:

1. Download NDIM-Engine-Windows-Portable.zip.
2. Extract the ZIP.
3. Open NDIM Engine Runtime/Launch NDIM Engine.cmd.

Data storage

The app stores local research data on the user's machine. In desktop mode the launcher creates folders for data, uploads, exports, logs, SDMX packages, backups, and support bundles.

Stress-test corpus

The stress-test corpus is included in the package. It contains route-specific CSV files plus a bonus TXT narrative for testing narrative ingestion, SDMX readiness, approval, encoding, modelling, and policy-output workflow.

Distribution

GitHub Releases is the recommended place to host the installer. Do not rely on local file paths as download links; upload NDIM Engine Setup.exe and NDIM-Engine-Windows-Portable.zip as release assets.

Known limitations

• This Windows build is unsigned, so Windows SmartScreen may warn users on first launch.
• Advanced Torch/Pyro analytics are not bundled in this package; deterministic fallback analytics remain available.
• macOS and Linux packages must be built on those operating systems.