Tom Eustace

20 years of shipping. Still building. Still learning.

[Vaishnavi @_vmlops]

Transformers Cheat Sheet

https://t.co/RBKYiSbnGG

[Tom Eustace @tomeustace]

Check out The Brand Design System by Open Session on @figma: https://t.co/vOBFcf2oOU

[Tom Eustace @tomeustace]

2026 Agentic Coding Trends Report.pdf https://t.co/Ez23RbWtqw

[Tom Eustace @tomeustace]

Looking forward to trying this out https://t.co/F8B7sLuHIy

[Vaishnavi @_vmlops]

Anthropic just launched its first certification: Claude Certified Architect

A 301-level exam for developers and architects building with the Claude ecosystem

Covers:
• Agentic Architecture & Orchestration
• Claude Code Workflows
• Tool Design & MCP Integration
• Prompt Engineering & Structured Output
• Context Management & Reliability

Free for Anthropic partners...Otherwise, $99

If you're already building with Claude, this is a solid way to validate your expertise

[Vaishnavi @_vmlops]

The CEO of Y Combinator, Garry Tan, just open-sourced his personal AI setup called GStack

It transforms Claude code into a fully functional virtual tech company, with distinct AI roles like CEO, engineering manager & QA tester

Github- https://t.co/ZzmRM5mQls

The best part? A conductor agent makes the AI think strategically before writing a single line of code letting you ship software like a YC founder

[Tom Eustace @tomeustace]

Worth a shot https://t.co/h3DJGyOFHZ

[Tom Eustace @tomeustace]

https://t.co/swFjfXXkX3

[Tom Eustace @tomeustace]

This looks like a powerful bit of kit https://t.co/guuxcENNW1

[Andrew Ng @AndrewYNg]

I'm excited to announce Context Hub, an open tool that gives your coding agent the up-to-date API documentation it needs. Install it and prompt your agent to use it to fetch curated docs via a simple CLI. (See image.)

Why this matters: Coding agents often use outdated APIs and hallucinate parameters. For example, when I ask Claude Code to call OpenAI's GPT-5.2, it uses the older chat completions API instead of the newer responses API, even though the newer one has been out for a year. Context Hub solves this.

Context Hub is also designed to get smarter over time. Agents can annotate docs with notes — if your agent discovers a workaround, it can save it and doesn't have to rediscover it next session. Longer term, we're building toward agents sharing what they learn with each other, so the whole community benefits.

Thanks Rohit Prsad and Xin Ye for working with me on this!

npm install -g @aisuite/chub

GitHub: https://t.co/OCkyxXQMCq

[Tom Eustace @tomeustace]

A comprehensive list of directories, communities, and platforms where indie developers can share their apps to get early users and feedback. https://t.co/xo0LjyeOYf

[Tom Eustace @tomeustace]

https://t.co/xo0LjyeOYf

[Mastra @mastra]

https://t.co/wCCwUhoWa3

[Shruti Codes @Shruti_0810]

I stopped paying $200/month for Claude Code.

And replaced it with… my laptop. 💻

No API bill.
No cloud dependency.
No data leaving my machine.

Here’s the 10-minute setup that changed everything 👇

---

⚡ Step 1: Install Ollama

code:
curl -fsSL https:// ollama .com/install.sh | sh

---

⚡ Step 2: Pull a coding model

code:
ollama pull qwen2.5-coder

You can also try:

•Qwen2.5-Coder** → Best all-around code generation
• DeepSeek-Coder** → Strongest at debugging & refactoring
• Llama 3** → Solid general reasoning

---

⚡ Step 3: Trick Claude Code

Point it to localhost:

code:
ANTHROPIC_BASE_URL=http://localhost:11434 claude

---

Now Claude Code thinks it’s talking to Anthropic.

It’s talking to your laptop. 😌

Let’s be real:

Local models ≠ Sonnet or Opus.
Massive multi-file refactors can struggle.
Long context windows get messy.

But for:
• Scaffolding
• Writing tests
• Refactoring small modules
• Boilerplate
• Quick edits

They’re more than enough.

And the benefits?

✅ $0/month
✅ 100% local
✅ Full privacy
✅ No API key risk

Your code never leaves your machine.
Your bill goes from $200 → $0.

This is the new indie dev stack.

🔖 Save this.
🔁 Repost to help someone cancel a subscription.
Follow for more unfair AI advantages. 🚀

[Tom Eustace @tomeustace]

@eliana_jordan Nice, congrats. I shipped my first app on Android but think ios might be a better bet.

[Hasan Toor @hasantoxr]

🚨This Python framework scrapes ANY website and extracts structured data in minutes.

It's called Scrapy and it crawls and extracts structured data from any website entirely from your own machine.

No SaaS scraping bills. No cloud API limits. No data leaving your infrastructure.

It's powered by Python's most battle-tested crawling engine with 59K stars and 15+ years of production use.

→ Define your spider once
→ Get clean structured data
→ Scale to millions of pages
→ Export to JSON, CSV, XML instantly

All running locally. Zero cloud dependency.

But it's not just a scraping script.

It's a full data extraction framework:

→ Async architecture for parallel crawling at scale
→ Built-in middleware for proxies, retries, and rate limiting
→ CSS and XPath selectors with zero boilerplate
→ Pluggable pipelines for cleaning, deduplication, and storage
→ 54,800+ production projects already depend on it

100% Opensource. BSD-3 License.

Works on macOS, Windows, and Linux right now.

This is the moment web scraping stops being a SaaS subscription and starts running on your own terms.

Link in the first comment 👇

[Tom Eustace @tomeustace]

@cheers_2_life87 @tom_doerr @grok I tried Docling which was very accurate but slow.

[Tom Dörr @tom_doerr]

Converts PDFs to Markdown at 100 pages per second on CPUs

https://t.co/2XTXTXaHrO

[Vaishnavi @_vmlops]

90% of developers still use Claude Code like a chatbot

That’s the old way

The real power is using it like an autonomous coding system

Here’s how top developers actually use Claude Code:
• /agents → Run specialized agents for different tasks
• /mcp → Connect Claude to tools, APIs, and databases
• /hooks → Automate workflows
• /sandbox → Test code safely

[Tom Eustace @tomeustace]

Just launched my first cursor cloud agent via Slack. This a great feature and less risky that other options IMO.

Though @cursor_ai I had to enable on demand usage to use cloud agents but can't set spend alerts as shown in docs. Is this a bug or am I missing somehting?