The difference between mcp/cli/api has been discussed at length but in general mcp gives agents oauth (for clients like Claude web/chatgpt) and a “more agentic” interface that can helpfully steer the agent instead of having it thrash tokens on erroneous api calls and request formats.

There are good and bad parts to both. Remote mcp is pretty good these days. The generative ui (mcp app) stuff is interesting and useful, if still immature.

Use what works for you!

I built Basic Memory — it imports your ChatGPT export and turns it into plain Markdown files. Every conversation becomes a file you can actually read, search, and use with whatever AI you switch to.

This is not an ad. It is free and open source. Your data belongs to you. Keep it.

Steps:

1. Settings → Data Controls → Export Data (ChatGPT emails you a zip)

2. Install Basic Memory (brew tap basicmachines-co/basic-memory && brew install basic-memory)

bm import chatgpt conversations.zip

Complete docs: http://docs.basicmemory.com

Local: https://github.com/basicmachines-co/basic-memory or hosted: http://basicmemory.com

Remote: Yes (US-based preferred)

Willing to relocate: No

Technologies: Python, FastAPI, Rust, JavaScript/TypeScript, React, Vue, Nuxt, AWS, GCP, Kubernetes, PostgreSQL, Redis, Terraform, Docker, Event Sourcing, CI/CD, LLM Integration

Résumé/CV: https://docs.google.com/document/d/1neImwQawfg8lbedopWElJFoU...

Email: paul@basicmachines.co

Senior/Staff/Principal Software Engineer with 20+ years building scalable SaaS platforms. Recently founded Basic Memory LLC, growing it from concept to 2,000+ GitHub stars and 56 paying customers in under 6 months, building one of the first production Model Context Protocol (MCP) implementations. Previously: Engineering Manager/Principal at companies like Fracture (50% AWS cost reduction), Spot Insurance, Pensa (ML/computer vision), and Spanning (100M+ events daily). Strong track record of rapid iteration, technical leadership, and delivering products that generate revenue. Looking for Staff/Principal Engineer or Engineering Manager roles at product companies building developer tools, SaaS platforms, or AI-powered products.

GitHub: github.com/basicmachines-co

Website: basicmemory.com

Senior/Staff/Principal Software Engineer with 20+ years building scalable SaaS platforms. Recently built Basic Memory (basicmemory.com) - an AI-powered knowledge management platform from concept to 2,000+ GitHub stars and 56 paying customers in 6 months.

What I do:

- Full-stack SaaS platform architecture and implementation

- Python/FastAPI backends, modern frontend (React/Vue/TypeScript)

- Multi-tenant cloud infrastructure (AWS/GCP, Kubernetes, Terraform)

- Event-driven systems, microservices, API design

- LLM integration and AI-powered workflows

- Technical leadership and team mentorship

Recent work:

- Built complete MCP (Model Context Protocol) server implementation - one of the first production deployments

- Multi-tenant SaaS with per-tenant isolation (Fly.io, PostgreSQL, Redis)

- OAuth authentication + billing infrastructure (Stripe, WorkOS)

- Semantic knowledge graph with full-text search

- Cross-platform sync architecture (web, desktop, mobile)

Available for:

- Architecture consulting and technical leadership

- Platform development (0->1 or scaling existing systems)

- AI/LLM integration projects

- Team augmentation for Python/cloud infrastructure work

Rate: $200-250/hour depending on project scope

Contact: paul@basicmachines.co | GitHub: github.com/basicmachines-co

What makes it different:

* Claude (and other MCP AIs) can both read from and write to your local markdown files

* Creates a connected network of notes that Claude can navigate between

* All data stays as markdown files on your computer

* When you chat with Claude, it automatically builds structure with simple patterns

* You can start a new chat and just say "Let's continue discussing X" without uploading files or copy/pasting

* Files are available locally, so you can edit them in and editor, like Obsidian, VS Code, etc.

I built this because I was tired of repeating myself to AI assistants and wanted my conversations to accumulate into something useful over time. It's basically a way to give Claude persistent memory while keeping your data local and be easily readable and editable my humans.

Some people are using it to maintain project context across sessions, document systems, and build topic maps without having to manually organize everything.

Demo video: https://basicmachines.co/images/Claude-Obsidian-Demo.mp4 GitHub: https://github.com/basicmachines-co/basic-memory

I built this because I found myself constantly repeating information to LLMs and wanted a system where my knowledge grew naturally through conversations while maintaining complete control over my data.

Demo video: https://basicmachines.co/images/Claude-Obsidian-Demo.mp4

Key features:

- Continue conversations exactly where you left off without repetition

- All knowledge stays in local Markdown files you can edit anytime

- Works with Claude Desktop via the Model Context Protocol

- Seamless integration with Obsidian for visualization and editing

- Fully open source (AGPL)

The system works by creating structure from simple markdown patterns:

- Observations with categories: `- [category] fact #tag`

- Relations between documents: `- relation_type [[WikiLink]]` or plain `[[Wikilinks]]`

- These patterns emerge naturally during conversations

When you chat with Claude, you can simply say "Let's continue our conversation about X" and it will build context from your knowledge base, without needing to upload files every time.

GitHub: https://github.com/basicmachines-co/basic-memory

Docs: https://memory.basicmachines.co/

Website: https://basicmachines.co/

Requires Claude Desktop or other MCP host and Python 3.12+

I'd love feedback from the HN community, particularly from those interested in knowledge management or AI applications.

Comments URL: https://news.ycombinator.com/item?id=43374258

Points: 4

# Comments: 0

In the past I've just used raw psql from the console, but having an IDE is a big quality of life improvement IMO.

Experienced developer, leader and creative problem solver, iteratively designing, delivering, and improving software architectures to deliver business value.

---

SKILLS:

- Expert guidance and practical solutions

- Tangible results without hype

- Extensive industry experience to address unique challenges effectively

---

SERVICES:

- Consulting Principal Engineer - Empowering Your Team for Success: hands-on mentoring, code review, task estimation and decomposition. Help build a team of 10x engineers, level up engineering processes, and foster a culture of continuous improvement and engineering craftsmanship.

- Architecture and Design Reviews - Ensuring a Strong Foundation: optimize your existing design and technology stack for maximum efficiency and reliability

- Cloud-Native Expertise - Optimize for Cloud: Transition smoothly to the cloud or optimize your existing cloud setup, minimize cloud costs, automate delivery pipelines, and leverage cloud-native architectures

- Implement MVP - From Idea to Launch: design systems that are modern, scalable, and maintainable. Iterate rapidly and fine-tune your product offering to deliver value quickly.

---

TECHNOLOGIES:

- Python, Java, Golang, Javascript/Typescript

- Linux, AWS, GCP, Kubernetes, Docker

- Microservices, Event Sourcing, Cloud Architecture, SaaS, Machine Learning (ML)

- PostgreSQL, Kafka

---

Location: US

Remote: Yes (Worldwide)

Email: paul [at] basicmachines.co

Comments URL: https://news.ycombinator.com/item?id=36951875

Points: 5

# Comments: 7

I’ve been interested in reading more about this. Do you have any pointers to good info?

I had a friend that used to do ironic yelp reviews. His review of the hot dogs at the 7-11 by his house were priceless.