mind

The Problem

You: "Remember that auth bug we fixed?"
Claude: "I don't have memory of previous conversations."
You: "We spent 3 hours on it yesterday"
Claude: "I'd be happy to help debug from scratch!"

200K context window. Zero memory between sessions.

You're paying for a goldfish with a PhD.

The Fix

You: "What did we decide about auth?"
Claude: "We chose JWT over sessions for your microservices.
        The refresh token issue - here's exactly what we fixed..."

One file. Claude remembers everything.

Installation

Claude Code

# One-time setup (if you haven't used GitHub plugins before)
git config --global url."https://github.com/".insteadOf "[email protected]:"

# In Claude Code
/plugin add marketplace memvid/claude-brain

Then: /plugins → Installed → mind Enable Plugin → Restart.

OpenCode

Step 1: Create opencode.json (if needed)

Check if your project has an opencode.json file. If not, create one in your project root:

{
  "$schema": "https://opencode.ai/config.json",
  "plugin": []
}

Step 2: Add claude-brain plugin

Add claude-brain to the plugin array:

{
  "$schema": "https://opencode.ai/config.json",
  "plugin": ["claude-brain"]
}

If you already have plugins, append to the array:

{
  "plugin": ["your-existing-plugin", "claude-brain"]
}

Step 3: Install the package

npm install claude-brain

Step 4: Restart OpenCode

Restart OpenCode to load the plugin.

Step 5: Verify installation

Run this command to confirm it works:

/mind stats

You should see memory statistics. If you see "Memory will be created", the plugin is working and will create the memory file on first use.

One-line Installer (Both Platforms)

curl -fsSL https://raw.githubusercontent.com/memvid/claude-brain/main/install.sh | bash

Done.

How it Works

After install, Claude's memory lives in one file:

your-project/
└── .claude/
    └── mind.mv2   # Claude's brain. That's it.

No database. No cloud. No API keys.

What gets captured:

• Session context, decisions, bugs, solutions
• Auto-injected at session start
• Searchable anytime

Why one file?

• git commit → version control Claude's brain
• scp → transfer anywhere
• Send to teammate → instant onboarding

Commands

In Claude Code or OpenCode:

/mind stats                       # memory statistics
/mind search "authentication"     # find past context
/mind ask "why did we choose X?"  # ask your memory
/mind recent                      # what happened lately

Or just ask naturally: "mind stats", "search my memory for auth bugs", etc.

Platform Comparison

Feature	Claude Code	OpenCode
Installation	/plugin add marketplace	opencode.json
Memory file	.claude/mind.mv2	.claude/mind.mv2
Commands	/mind:*	/mind:*
Skills	skills/ directory	skills/ directory
Observation capture	PostToolUse hook	tool.execute.after
Session context	SessionStart hook	session.created event

CLI (Optional)

For power users who want direct access to their memory file:

npm install -g memvid-cli

memvid stats .claude/mind.mv2           # view memory stats
memvid find .claude/mind.mv2 "auth"     # search memories
memvid ask .claude/mind.mv2 "why JWT?"  # ask questions
memvid timeline .claude/mind.mv2        # view timeline

Full CLI reference →

FAQ

How big is the file?

Empty: ~70KB. Grows ~1KB per memory. A year of use stays under 5MB.

Is it private?

100% local. Nothing leaves your machine. Ever.

How fast?

Sub-millisecond. Native Rust core. Searches 10K+ memories in <1ms.

Reset memory?

rm .claude/mind.mv2

Can I use the same memory file for both Claude Code and OpenCode?

Yes! The .mv2 file is portable and works with both platforms. Share one memory across tools.