🎉 Context Sync v0.6.1 - Multi-Session Support + Big File Intelligence

Hey PH! 👋

We just shipped v0.6.1 with two major features that our community asked for:

1️⃣ True Multi-Session Architecture

Every AI platform gets its own independent instance!

The Problem: Context Sync previously had ONE global "current project" in the database. Switching projects in ANY platform would break ALL other sessions.

The Solution: Each MCP server instance now maintains its own session state in memory:

┌─────────────────┐  ┌─────────────────┐  ┌─────────────────┐
│  Claude Desktop │  │  Cursor IDE     │  │  VS Code Copilot│
│                 │  │                 │  │                 │
│  MCP Instance 1 │  │  MCP Instance 2 │  │  MCP Instance 3 │
│  Project A      │  │  Project B      │  │  Project A      │
└────────┬────────┘  └────────┬────────┘  └────────┬────────┘
         │                    │                    │
         └────────────────────┼────────────────────┘
                              ▼
                    Shared Database
              (decisions, todos, context)

What this means:

• Work on multiple projects simultaneously across ANY AI platform

• Each tool (Claude, Cursor, Copilot, etc.) has its own independent session

• Switch projects mid-conversation without affecting other tools

• All platforms still share the same project data

2️⃣ Smart File Skimming for Large Codebases

Big files matter too - Context Sync now handles them intelligently!

The Problem: Large codebases have massive files (bundle outputs, generated code, data files). Previously, we'd either skip them entirely or choke on them.

The Solution: Intelligent file skimming for files over 1MB:

• ✅ Read full header (first 32KB) - where imports/definitions live

• ✅ Read full footer (last 16KB) - where exports/conclusions are

• ✅ Pattern-match through middle for key code (functions, classes, imports)

• ✅ Strategic sampling of content

What this means:

• Big codebases with 1000+ files? No problem!

• Generated files, bundles, large configs? We handle them!

• 80%+ reduction in data processed, 100% of the context retained

• Your AI assistant now sees the FULL picture of your codebase

💡 Real-World Impact

Multi-Session Architecture:

Scenario: You're working on 3 projects
- Claude Desktop: Debugging Project A (microservices)
- Cursor: Building features for Project B (monorepo)
- VS Code: Reviewing Project A code

Before v0.6.1: Chaos. Sessions conflicting.
After v0.6.1: Perfect harmony. Each tool independent!

Big File Intelligence:

Scenario: React app with huge bundle files
- dist/bundle.js (5MB) - Previously ignored
- Now: Skimmed intelligently, imports/exports captured
- Result: AI knows your full dependency graph

Together: Work on massive projects across multiple AI tools without compromise!

🚀 Get Started

bash

npm install -g @context-sync/server

⭐ GitHub: github.com/Intina47/context-sync

This was one of our most requested features - built based on YOUR feedback!

What should we build next? 🙌