Features Overview

Discover all the powerful features that make Mini-CoderBrain the ultimate context awareness system for Claude Code.


🎯 Core Features

1. Perfect Context Awareness

Problem: Claude forgets your project structure, tech stack, and decisions between conversations.

Solution: Mini-CoderBrain loads complete project context once per session and maintains it naturally.

Benefits:

  • ✅ Zero repeated explanations
  • ✅ Claude knows your tech stack
  • ✅ Remembers architectural decisions
  • ✅ Understands your coding patterns
  • ✅ Perfect continuity across sessions

How it works:

Session Start → Loads 5 memory files → Context persists in conversation history

2. Real-Time Activity Tracking

Feature: Live operation counter that shows exactly what’s happening.

Display:

📊 Activity: 42 ops | 🗺️ Map: Loaded | ⚡ Context: Active

Tracks:

  • Read operations
  • Write operations
  • Edit operations
  • Grep searches
  • Glob searches
  • Bash commands
  • File operations

Benefits:

  • See exactly what Claude is doing
  • Understand development velocity
  • Track productivity metrics
  • Identify patterns in work

3. Intelligent Notifications

Feature: Proactive suggestions based on session activity and memory state.

Notifications:

  • 🧹 Memory Cleanup: “Consider running /memory-cleanup (activeContext.md is 215 lines)”
  • 🗺️ Map Rebuild: “15 new files detected. Run /map-codebase –rebuild”
  • 💾 Memory Sync: “High activity session (47 ops). Consider /memory-sync”
  • 📊 Progress Update: “Completed 5 tasks. Update progress.md?”

Smart Triggers:

  • File count thresholds
  • Memory file size limits
  • Operation count tracking
  • Session duration

4. Zero Context Duplication

Problem: Old versions injected context repeatedly, causing “Prompt is too long” errors.

Solution: Context loaded once at session start, persists naturally in conversation history.

Impact:

  • 79.9% token reduction vs. previous versions
  • 25% longer conversations before limits
  • Zero duplication - context loaded once
  • Perfect continuity - full awareness maintained

Technical:

Turn 1: Load 782 lines of context → Enter conversation history
Turn 2+: Context already available → No re-injection needed
Result: Natural conversation, perfect awareness, minimal tokens

5. Automatic Memory Management

Feature: Prevents “Prompt is too long” errors through intelligent cleanup.

How it works:

  1. Session activity tracking counts operations
  2. Smart notifications suggest cleanup when needed
  3. /memory-cleanup archives old data
  4. Keeps last 5 session updates, archives rest
  5. No data loss - everything preserved in .claude/archive/

Benefits:

  • ✅ Never hit “Prompt is too long” again
  • ✅ ~60% context token reduction
  • ✅ All history preserved in archive
  • ✅ Automatic suggestions when needed

6. Codebase Mapping System

Feature: Revolutionary instant file access system.

Usage:

/map-codebase --rebuild    # One-time setup (~30 seconds)
/map-codebase              # Instant loading (future sessions)

What it does:

  • Scans entire codebase
  • Creates semantic file map
  • Enables instant file location
  • Surgical precision development

Benefits:

  • Claude knows EXACT file locations
  • No more “Where is the User model?” questions
  • Instant file access
  • Perfect for large codebases

Example:

User: "Update the User model"
Claude: *Checks codebase-map.json → finds src/models/User.ts*
Claude: *Opens file and makes changes, no questions asked*

7. Mandatory Pre-Response Protocol

Feature: Claude MUST check context before responding.

Protocol:

Before ANY response:
1. ✅ CHECK productContext.md → Project info
2. ✅ CHECK systemPatterns.md → Coding patterns
3. ✅ CHECK activeContext.md → Current focus
4. ✅ CHECK project-structure.json → File locations
5. ✅ CHECK codebase-map.json → Specific files

ONLY AFTER checking → Respond to user

Banned Questions:

  • ❌ “What framework are you using?” → CHECK productContext.md
  • ❌ “Where is the User model?” → CHECK codebase-map.json
  • ❌ “What’s your coding style?” → CHECK systemPatterns.md
  • ❌ “Should I create tests?” → CHECK systemPatterns.md

Result: Claude uses context automatically, no unnecessary questions!


8. Cross-Session Memory

Feature: Perfect continuity between sessions.

How it works:

Session End:
- Stop hook saves session summary
- Updates activeContext.md
- Preserves progress and decisions

Session Start:
- Session-start hook loads context
- Picks up exactly where you left off
- Full awareness of past work

Benefits:

  • ✅ Resume work instantly
  • ✅ No context loss between sessions
  • ✅ Remember all decisions
  • ✅ Track progress over days/weeks

9. Universal Project Detection

Feature: Automatically detects your project type and structure.

Detects:

  • Frontend paths (src/, app/, components/)
  • Backend paths (server/, api/, routes/)
  • Database paths (prisma/, migrations/, models/)
  • Test paths (tests/, tests/, spec/)
  • Config files (package.json, requirements.txt, Cargo.toml)

Supported:

  • JavaScript/TypeScript (React, Next.js, Node, etc.)
  • Python (Django, Flask, FastAPI, etc.)
  • Rust, Go, Java, C++, PHP, Ruby
  • Mobile (React Native, Flutter, Swift, Kotlin)
  • ANY project type!

10. Memory Bank System

5 Core Memory Files:

productContext.md

  • Project name and overview
  • Technology stack
  • Architecture
  • Core features
  • When to edit: Once during setup

activeContext.md

  • Current focus
  • Recent achievements
  • Next priorities
  • Active blockers
  • When to edit: Auto-updated by hooks

progress.md

  • Sprint tracking
  • Completed tasks
  • In-progress work
  • Upcoming priorities
  • When to edit: Auto-updated

decisionLog.md

  • Technical decisions (ADRs)
  • Timestamped entries
  • Rationale and impact
  • When to edit: Auto-appended

systemPatterns.md

  • Coding standards
  • Testing approach
  • Error handling
  • Security patterns
  • When to edit: Once during setup

🛠️ Advanced Features

Real-Time Context Updates

Quick updates without full sync:

/context-update focus "Building authentication"
/context-update blocker "API rate limiting issue"
/context-update achievement "Completed user registration"
/context-update priority "Add email verification"

Smart Memory Sync

Full context synchronization:

/memory-sync          # Smart sync based on activity
/memory-sync --full   # Full sync (all files)
/memory-sync --quick  # Quick sync (activeContext only)

Manual Quick Sync

Fast sync with note:

/umb "Completed authentication feature"

Memory Cleanup

Archive old data:

/memory-cleanup           # Normal cleanup (keep last 5)
/memory-cleanup --dry-run # Preview changes
/memory-cleanup --full    # Aggressive (keep last 3)

Every response includes:

🧠 MINI-CODERBRAIN STATUS
📊 Activity: 42 ops | 🗺️ Map: Loaded | ⚡ Context: Active

💡 Consider running /memory-cleanup (activeContext.md is 215 lines)

Components:

  • Activity: Operation count this session
  • Map: Codebase map status (None/Loaded/Available)
  • Context: Context awareness status
  • Notifications: Smart suggestions (only when relevant)

🎯 Use Cases

Solo Developers

  • Track feature development
  • Remember architectural decisions
  • Maintain coding standards
  • Monitor progress

Teams

  • Share project context
  • Document decisions automatically
  • Maintain consistency
  • Onboard new members

Learning

  • Document learning progress
  • Track concepts mastered
  • Build knowledge base
  • Maintain study goals

Open Source

  • Document contribution patterns
  • Track community decisions
  • Maintain project history
  • Onboard contributors

🚀 Performance Benefits

Benefit Impact
Context Duplication Eliminated (0%)
Token Efficiency 79.9% reduction
Conversation Length 25% longer
Memory Bloat Prevented automatically
Setup Time 30 seconds
Learning Curve Zero
Configuration None required

🔐 Security Features

Gitignore Protection

  • .claude/memory/*.md files gitignored
  • User data never committed
  • Framework files committed
  • Templates committed (not actual data)

Privacy

  • All context stored locally
  • No external services
  • No data transmission
  • Full control over sensitive info

🌟 What Makes It Special

1. Zero Configuration

Drop in and go. No config files, no setup, no learning curve.

2. Universal Compatibility

Works with ANY project type, ANY language, ANY framework.

3. Automatic Everything

  • Auto-detects project structure
  • Auto-tracks activity
  • Auto-updates context
  • Auto-suggests maintenance

4. Perfect Continuity

Never lose context between sessions. Ever.

5. Production Ready

Built for real-world development, tested in production.


Ready to see the commands?

View Commands Reference →