Skip to content

QUICK_START_GUIDE.md

Latest commit

 

History

History
134 lines (101 loc) · 4.25 KB

QUICK_START_GUIDE.md

File metadata and controls

134 lines (101 loc) · 4.25 KB

Quick Start Guide

Get your AI to remember your project context in 5 minutes.

tinyMem is a local tool that sits between your code and your AI assistant. It creates a "memory brain" for your specific project so you don't have to keep repeating context.

1. Install tinyMem

First, get the single executable file. No complex installers or dependencies required.

Windows

  1. Download the latest release (tinymem-windows-amd64.exe).
  2. Create a folder C:\Tools (or use an existing one) and put the file there.
  3. Rename it to tinymem.exe for convenience.
  4. Important: Add this folder to your PATH so you can run it from anywhere.
    • Search "Edit the system environment variables" > "Environment Variables" > Select "Path" in User variables > "Edit" > "New" > Paste C:\Tools > OK > OK.

macOS / Linux

  1. Open your terminal.
  2. Run this command to download and install to /usr/local/bin: 
    os="$(uname -s | tr '[:upper:]' '[:lower:]')"
arch="$(uname -m)"
case "$arch" in
  x86_64|amd64) arch="amd64" ;;
  aarch64|arm64) arch="arm64" ;;
  *) echo "Unsupported arch: $arch" >&2; exit 1 ;;
esac
curl -L "https://github.com/daverage/tinyMem/releases/latest/download/tinymem-${os}-${arch}" -o tinymem
chmod +x tinymem
sudo mv tinymem /usr/local/bin/

2. Initialize Your Project

tinyMem creates a memory database inside your project folder (in a hidden .tinyMem directory). You need to tell it which project to manage.

  1. Open your terminal.
  2. Navigate to your project's root folder: 
    cd /path/to/my-cool-app
  3. Initialize the memory system: 
    tinymem health
    (You should see "✅ System is healthy" and a new .tinyMem folder created).

3. Connect Your AI

Choose the method that matches how you work.

Option A: IDE Integration (Claude Desktop, Cursor, VS Code)

Best for: Coding assistants and chat interfaces.

You need to tell your IDE to run tinyMem as a "Model Context Protocol" (MCP) server.

For Claude Desktop: Edit your config file (usually ~/Library/Application Support/Claude/claude_desktop_config.json on macOS or %APPDATA%\Claude\claude_desktop_config.json on Windows):

{
  "mcpServers": {
    "tinymem": {
      "command": "tinymem",
      "args": ["mcp"]
    }
  }
}

Restart Claude Desktop. The 🔌 icon should appear, indicating tinyMem is connected.

Option B: Proxy Mode (Scripts & API Clients)

Best for: Running Python scripts, Aider, generic OpenAI clients, or terminal tools.

  1. Configure for Local LLMs (Optional): If you use LM Studio or Ollama, create a file at .tinyMem/config.toml:

    [proxy]
base_url = "http://localhost:1234/v1" # Point to LM Studio

  2. Start the proxy in a separate terminal window:

    cd /path/to/my-cool-app
tinymem proxy

  3. In your main terminal, set the environment variable to route requests through tinyMem:

    export OPENAI_API_BASE_URL=http://localhost:8080/v1

    For Aider: aider --openai-api-base http://localhost:8080/v1 --model openai/qwen2.5-coder-7b-instruct

  4. Run your tool or script as usual. tinyMem will transparently intercept and inject memory.

4. Verify It's Working

  1. Ask your AI something about your project.
  2. Check the tinyMem status: 
    tinymem stats
  3. See your memories visually: 
    tinymem dashboard

🛠 Troubleshooting & Logging

If things aren't working as expected, you can enable full debug logging:

  1. Environment Variable: export TINYMEM_LOG_LEVEL=debug
  2. Config File: Set level = "debug" in the [logging] section of .tinyMem/config.toml.

Logs are stored by default in .tinyMem/logs/tinymem-<date>.log.

Next Steps

  • Read the full README for advanced configuration.
  • Learn about Chain-of-Verification (CoVe) to understand how tinyMem filters hallucinations automatically (enabled by default).
  • Learn about Memory Types to understand the difference between a fact (verified) and a claim (unverified).
  • Check tinymem doctor if you run into any issues.