Your first Markdown project,
inside AI OS.

Use the Current 1.1.1 download from the Builder Start Here post. Extract the ZIP into a new folder and keep the whole package together. This guide follows Current, not the separate Classic 1.0.4 design.

Open a terminal inside the extracted dashboard folder. On macOS, type cd , drag that folder from Finder into Terminal and press Return. On Windows, open the folder in File Explorer, type powershell in its address bar and press Enter.

python3 --version
python3 setup.py --check-only
python3 setup.py

py -3 --version
py -3 setup.py --check-only
py -3 setup.py

The version check must report 3.11 or newer. Verification should say Package verified. The launcher opens the dashboard or prints a local address beginning http://127.0.0.1:. Keep the terminal open while using it. Open that address, not index.html directly. Keep this server local; this walkthrough does not require port forwarding or a public tunnel.

Need Python? Install it from python.org, then reopen your terminal. If Windows has no py -3 command but python --version reports 3.11+, use python instead. Linux instructions are included in the package, but Linux installation has not been verified.

Download the free sample folder and extract it somewhere outside the dashboard folder. It contains only these three fictional Markdown notes:

sample-project/
├── projects/
│   └── launch/
│       └── brief.md
├── notes/
│   └── meeting.md
└── reference/
    └── checklist.md

The sample ZIP is free. It contains text files, not the paid dashboard application. Keep the extracted sample-project folder in place: the dashboard reads the original files when you open them.

# Fictional launch project

Goal: prepare a short product walkthrough.

Next: draft the outline and choose a demonstration.

Stop the running dashboard with Ctrl + C in its terminal. Stay in the dashboard folder. Run the command below, replacing the example path with the full path to your extracted sample-project folder. Keep the quotes around paths with spaces.

python3 setup.py --workspace "/full/path/to/sample-project"

py -3 setup.py --workspace 'C:\full\path\to\sample-project'

The dashboard opens in LOCAL FILES mode. This example produces 7 graph items and 6 links: one workspace root, three top-level folder groups and three files. Those are structural links, not AI-generated connections between ideas.

Already have an imported workspace? Use a separate clean extraction for this walkthrough. Reimporting replaces the file index in that package; it does not merge two workspaces. Your source notes remain in their original folder.

In the graph toolbar, click search and type brief. Select the result named Brief with path projects/launch/brief.md. This search finds indexed file information; it is not an AI question-answering box.

Its file card appears. Click OPEN to read the text. You should see the heading # Fictional launch project and the walkthrough goal below it.

To reopen the same imported index later, run this from the dashboard folder:

python3 setup.py --user

py -3 setup.py --user

When you add, rename or remove source files, stop the server and repeat the --workspace command with the same folder path. Reopen or refresh the browser. The index is manual, not continuously synchronised. Existing custom panels are preserved during reimport and may still show their previous counters. Check newly imported files through the graph and search.

The importer accepts up to 2,000 UTF-8 Markdown/text files, each at most 256 KB. Start small. The chosen source path is stored in the dashboard's data/user/source.json, so keep an imported package private. Do not post your notes, paths or credentials in support screenshots.

If you already use MAPS, you can explicitly choose that folder once the sample works. Use this package's existing interface: the free MAPS guide's Cards 4 and 5 build a screen from scratch and should not be applied over your downloaded dashboard.

A FEW QUICK CHECKS

If something looks wrong.

Checked on 10 November 2026 using the published Current 1.1.1 archive: package verification, three-file import, browser search and OPEN. Source-file hashes remained unchanged. Screenshots use fictional data. The release also has prior Windows x64 VM checks; this walkthrough was rerun on macOS.