🧠 claude-brain

A persistent memory and workflow system for Claude Code.
Custom slash commands, a structured memory layer, and a mission-control cockpit —
so Claude knows who you are, what you're building, and what to do next from message one.

---

🤔 The problem

Every new Claude session starts cold. Claude doesn't remember your projects, your preferences, your open threads, or how you like to work. You waste the first 10 messages re-explaining context you already gave last time.

claude-brain fixes that.

---

✨ What's included

4 custom slash commands

Command	What it does
/standup	Morning orientation — active projects, open threads, suggested next move
/capture <thought>	Dump any idea or task and Claude routes it to the right project/note automatically
/ship <thing>	Pre-publish checklist for content (hook, thumbnail, CTA) or code (runs? secrets? mode?)
/brain	Claude audits its own memory — finds stale, missing, or duplicate notes and fixes them

Structured memory layer

~/.claude/projects/<project>/memory/
├── MEMORY.md              ← index (auto-loaded every session)
├── 00-operating-brief.md  ← READ FIRST: who you are, what's active, standing rules
├── how-to-work-with-user.md ← communication style and guardrails
└── [project notes].md     ← one file per project/topic

Claude reads MEMORY.md at the start of every session — so your context is always loaded without you having to repeat yourself.

Mission Control cockpit

A single MISSION-CONTROL.md file at your workspace root — one-glance view of:

• All active projects + current state + next move
• Open threads needing your action
• Protected items Claude should never touch

---

🚀 Install

# 1. Clone
git clone https://github.com/highqualityhq/claude-brain.git
cd claude-brain

# 2. Run the installer
chmod +x install.sh
./install.sh

Or manually:

# Copy commands
cp commands/*.md ~/.claude/commands/

# Copy memory templates to your project
mkdir -p ~/.claude/projects/<your-project>/memory/
cp memory/*.md ~/.claude/projects/<your-project>/memory/

---

⚙️ Setup (5 minutes)

Step 1 — Fill in the operating brief

Open ~/.claude/projects/<your-project>/memory/00-operating-brief.md and replace the template placeholders with your real info:

• Who you are and your goals
• Your active projects
• Standing rules (what Claude should never touch)
• Where your files live

Step 2 — Create your Mission Control

Copy the template and fill it in:

cp examples/MISSION-CONTROL.md ~/Projects/MISSION-CONTROL.md

Step 3 — Test it

Open a new Claude Code session and type:

/standup

Claude should immediately give you a briefing based on your operating brief. If it does — you're set.

---

📁 File structure

claude-brain/
├── commands/
│   ├── standup.md       # /standup command
│   ├── capture.md       # /capture command
│   ├── ship.md          # /ship command
│   └── brain.md         # /brain command
├── memory/
│   ├── MEMORY.md                  # index template
│   ├── 00-operating-brief.md      # session brief template
│   └── how-to-work-with-user.md   # communication style template
├── install.sh           # one-command installer
└── README.md

---

💡 How the memory system works

Claude Code automatically loads MEMORY.md from your project's memory directory at the start of each session. Each line in MEMORY.md points to a topic file — Claude reads the ones relevant to the current task.

The 00-operating-brief.md is pinned at the top with a ⭐ so Claude reads it first every session. It's your permanent context layer — update it whenever a project ships or a fact changes.

The /brain command audits the whole system periodically: finds stale info, broken links, and missing notes, then proposes fixes.

---

🧩 Recommended workflow

Morning  →  /standup          (orient, pick a thread)
Build    →  /capture <idea>   (never lose a thought)
Ship     →  /ship <thing>     (pre-publish checklist)
Weekly   →  /brain            (keep the memory sharp)

---

📄 License

MIT © highqualityhq