HackerTrans
TopNewTrendsCommentsPastAskShowJobs

mambacodes

no profile record

Submissions

[untitled]

1 points·by mambacodes·4개월 전·0 comments

[untitled]

1 points·by mambacodes·5개월 전·0 comments

[untitled]

1 points·by mambacodes·5개월 전·0 comments

[untitled]

1 points·by mambacodes·5개월 전·0 comments

[untitled]

1 points·by mambacodes·7개월 전·0 comments

[untitled]

1 points·by mambacodes·7개월 전·0 comments

[untitled]

1 points·by mambacodes·7개월 전·0 comments

[untitled]

1 points·by mambacodes·8개월 전·0 comments

[untitled]

1 points·by mambacodes·8개월 전·0 comments

[untitled]

1 points·by mambacodes·8개월 전·0 comments

[untitled]

1 points·by mambacodes·8개월 전·0 comments

[untitled]

1 points·by mambacodes·9개월 전·0 comments

[untitled]

1 points·by mambacodes·9개월 전·0 comments

[untitled]

1 points·by mambacodes·9개월 전·0 comments

Start coding in Cursor, finish in Claude – full context follows you

github.com
5 points·by mambacodes·9개월 전·6 comments

Cross-Platform Al Sync (Claude Desktop <> Cursor IDE)

github.com
3 points·by mambacodes·9개월 전·2 comments

comments

mambacodes
·5개월 전·discuss
You send events (`ingest`), fetch relevant context (`retrieve`), and feed quality signals (`feedback`). Orbit handles:

- memory storage and ranking - decay and compression - adaptive personalization (inferred profile memories) - API runtime concerns (auth, quotas, rate limits, observability)## Integration Modes

You can integrate Orbit in three ways.

| Mode | Use when | Entry point | | --- | --- | --- | | Python SDK | Your app is Python and you want fastest integration | `from orbit import MemoryEngine` | | REST API | Non-Python stack or service-to-service architecture | `POST /v1/ingest`, `GET /v1/retrieve`, `POST /v1/feedback` | | OpenClaw Plugin | You run OpenClaw agents and want memory wiring without custom glue code | `integrations/openclaw-memory/` |
mambacodes
·7개월 전·discuss
Ask AI to generate feature docs → AI writes them to Notion. Pull project specs from Notion → AI understands your plan instantly.

• "Create API documentation in Notion" • "Export this architecture decision to Notion" • "Read the product spec from Notion and implement it"

Context Sync is the open-source infrastructure that gives AI systems persistent memory across all your development tools, sessions, and projects.

Try it out: https://github.com/Intina47/context-sync
mambacodes
·8개월 전·discuss
- Universal AI Platform Registry supporting 8+ platforms - Smart Database Migration with 35-60% performance boost - Intelligent Context Analysis for auto-extraction - Enhanced Platform Detection and setup guides - Comprehensive documentation and migration tools

BREAKING CHANGE: Universal platform support with enhanced type system
mambacodes
·8개월 전·discuss


  Hey HN! 

  I built HostScan after getting tired of manually typing IP addresses like
  http://192.168.1.68:3000 into my phone every time I wanted to test responsive designs on real
  devices.

  What it does:
  - Click one button in VS Code
  - Automatically detects your running dev server (Next.js, Vite, Live Server, etc.)
  - Generates a QR code
  - Scan with your phone → instant mobile testing

  Key features:
  • Auto port detection (scans for ports 3000, 5173, 5500, etc.)
  • Smart IP filtering (ignores Docker/WSL/Hyper-V virtual adapters)
  • 100% local network (no external services, no accounts)
  • Works with Next.js, Vite, React, Angular, Python, and more

  It's open source (MIT): https://github.com/Intina47/hostscan

  The biggest gotcha I learned: Most dev servers only listen on localhost by default. Users need
  to bind to 0.0.0.0 (like `next dev -H 0.0.0.0`). The README has a comprehensive guide on this.

  Would love feedback from the community! What other pain points do you have with mobile testing?
mambacodes
·9개월 전·discuss
Hi HN!

I've been working on Context Sync (https://github.com/Intina47/context-sync) - an MCP server that gives AI coding assistants persistent memory across platforms and sessions.

Today I'm excited to share v0.6.0, which adds VS Code/GitHub Copilot support and major performance optimizations.

## The Problem We've all been there: you're deep into a coding session with Claude, hit the context limit, start a new chat, and suddenly Claude has forgotten your entire project architecture. You spend 30 minutes re-explaining the same context. Every. Single. Day.

## The Solution Context Sync gives AI assistants persistent memory. Start a project Monday in Cursor, continue Tuesday in Claude Desktop, debug Wednesday in VS Code - the AI remembers everything.

## What's New in v0.6.0 • *VS Code + GitHub Copilot support* - Now works across all major AI coding platforms • *2-5x faster database operations* - Prepared statement caching • *File size safety limits* - Prevents OOM crashes with 5MB max file size • *Real-time cache invalidation* - File watchers automatically update caches • *Async file operations* - Non-blocking I/O for better performance

## Technical Details Built as an MCP (Model Context Protocol) server in TypeScript with SQLite. Local-first, no cloud sync required. Uses intelligent caching, file watching with chokidar, and async/await throughout.

The performance optimizations were crucial - we were seeing memory issues with large codebases, so v0.6.0 adds file size limits, async operations, and much smarter caching.

## Try It ```bash

npm install -g @context-sync/server
mambacodes
·9개월 전·discuss
Thanks for sharing, I will check it out. We just launched our solution in PH today if you have a minute ab upvote would mean the world

https://www.producthunt.com/posts/context-sync
mambacodes
·9개월 전·discuss
Launched Context Sync on Product Hunt this morning (7:30 AM).

12 hours later: #17 with 98 upvotes. Not top 10, but not bad for a first launch.

What went right: - Strong organic validation (68 upvotes before tapping network) - Amazing feedback from Reddit/HN community - Clear product-market fit signals - Technical discussions were high quality

What went wrong: - Product Hunt bug: couldn't see ANY comments for 6+ hours - Messaging too technical (MCP jargon confused people) - WhatsApp groups barely converted (80→82 upvotes) - Didn't hit critical mass for algorithm boost

Best feedback received: - "Why isn't this built into Claude?" (fair question - business incentives) - VS Code support most requested (on roadmap) - Comparisons to Traycer, basic-memory (researching now) - Mobile view wanted (waiting for MCP support)

What I learned: 1. Lead with problem, not solution (people don't know what MCP is) 2. Test comment visibility EARLY in launch 3. Plain English > technical jargon 4. Reddit engagement > WhatsApp conversion 5. First launch is about learning, not winning

Next launch (v0.6.0): Ironically, we're building launch automation. The irony of manually launching a launch automation tool is not lost on me.

Goal: Top 10 on Product Hunt.

For other first-time launchers: - Your first launch will teach you more than any guide - Engagement matters more than upvotes - Community feedback is gold - Perfect is the enemy of shipped

Context Sync: https://github.com/Intina47/context-sync Product Hunt: https://www.producthunt.com/posts/context-sync

Building in public. What would you do differently?
mambacodes
·9개월 전·discuss
Hey HN!

We just launched Context Sync on Product Hunt and would love your support: https://www.producthunt.com/products/context-sync-local-mcp-...

Quick recap for those who missed our last post:

Context Sync solves the "explain my project for the 100th time" problem. It's an MCP server that gives Claude persistent memory across all conversations.

The magic moment: - Build in Cursor, hit context limit - Switch to Claude Desktop - Continue exactly where you left off - Zero re-explaining

What's included: - Cross-platform context sync (Cursor ↔ Claude) - Global TODO list with smart filtering - Code analysis (dependencies, call graphs, types) - Git integration - 50+ MCP tools for file operations - 100% local (SQLite) - your data never leaves

v0.5.0 shipped last week with the TODO system. Now you can: - Create tasks in Cursor - Check them in Claude - Mark complete anywhere - Everything syncs

The PH launch has been awesome so far.

If you've been frustrated with AI context switching, we built this for you.

Product Hunt: https://www.producthunt.com/products/context-sync-local-mcp-... GitHub: https://github.com/Intina47/context-sync

Setup takes <30 seconds. MIT licensed.

Would mean a lot if you could check it out and leave feedback!
mambacodes
·9개월 전·discuss
LOL right? I think we all rage-quit Cursor at the same context limit wall and decided to fix it.

What angle are you taking? I'm doing MCP + local SQLite. Betting you have some insights I'm missing.

Drop a link if you're sharing - would love to compare approaches. This space needs more collaboration.
mambacodes
·9개월 전·discuss
Nice work! The autowired routing + ActiveRecord ORM combo is clean.

Quick question: How does the AI runtime part work? Is it for code generation or something else?

We just shipped Context Sync (MCP server for persistent AI context). If you're doing AI-assisted PHP dev, might be interesting to connect Onetone with it - your framework decisions could persist across AI conversations.

Anyway, good luck with the launch! PHP frameworks don't get enough love these days.
mambacodes
·9개월 전·discuss
30 sec setup:

npm install -g @context-sync/server@latest
mambacodes
·9개월 전·discuss
Hey YC!

You know that moment when you're deep in Cursor, building something, hit the context window limit, and realize you'd have to explain EVERYTHING again to Claude from scratch?

Yeah, that's gone now.

Context Sync v0.5.0 lets you literally switch between Cursor and Claude mid-project and pick up exactly where you left off.

Real scenario I had yesterday: 1. Building a TODO app in Cursor 2. Hit context limit at 200K tokens 3. Open Claude Desktop 4. "Continue building the TODO app" 5. Claude: "Sure! Let me add the overdue detection to the TodoManager class we were working on..."

Claude knew: - The exact app I was building - The architecture decisions we made - The TODOs I created while coding - Every file structure and function

It's like having one continuous conversation across tools.

How it works: - MCP server stores your project context (decisions, conversations, TODOs) - Both Cursor and Claude connect to the same MCP server - Context auto-loads when you switch platforms - SQLite storage - everything local, nothing cloud

New in v0.5.0 - Global TODO System: - Create todos in Cursor: "todo: refactor auth, high priority" - Switch to Claude: "show my high priority todos" - Complete them anywhere: "mark auth refactor as done" - They persist across EVERYTHING

The pain this solves: - No more "let me explain my project again" - No more copy-pasting context between tools - No more losing track of decisions when switching - No more "which AI tool has the latest info?"

Technical: - 50+ MCP tools (file ops, git, code analysis, TODOs) - TypeScript + SQLite - Works with Claude Desktop + Cursor (VS Code coming) - MIT licensed

Setup: 2 minutes ```bash git clone https://github.com/Intina47/context-sync cd context-sync node setup.js ```

The moment you hit a context limit and realize you can just... switch tools and keep going? That's the magic.

Questions welcome!
mambacodes
·9개월 전·discuss
Stop copy-pasting context or re-explaining your project every time you switch tools. Work in Claude Desktop for brainstorming, jump to Cursor for coding, and pick up exactly where you left off - with full memory of what you discussed and decided. The first MCP server to enable true multi-platform AI collaboration. Your context follows you, not the other way around.