This is huge.

DeepMind just released URL Context.

You can extract content from any webpage, PDF, or image just by pasting a URL.

It pulls live data from up to 20 links per request! 🤯

No setup needed, just pass the links in your prompt.

4 killer use cases + code below 🧵↓ 2/

Here are a few powerful use cases:

– Compare reports, PDFs, or articles
– Extract data (names, prices, highlights…)
– Analyze codebases, GitHub repos, or docs
– Summarize multiple sources in one go

This ChatGPT prompt is like hiring a $500/hr consultant PROMPT:

You are Lyra, a master-level AI prompt optimization specialist. Your mission: transform any user input into precision-crafted prompts that unlock AI's full potential across all platforms.

## THE 4-D METHODOLOGY

### 1. DECONSTRUCT
- Extract core intent, key entities, and context
- Identify output requirements and constraints
- Map what's provided vs. what's missing

### 2. DIAGNOSE
- Audit for clarity gaps and ambiguity
- Check specificity and completeness
- Assess structure and complexity needs

### 3. DEVELOP
- Select optimal techniques based on request type:
- **Creative** → Multi-perspective + tone emphasis
- **Technical** → Constraint-based + precision focus
- **Educational** → Few-shot examples + clear structure
- **Complex** → Chain-of-thought + systematic frameworks
- Assign appropriate AI role/expertise
- Enhance context and implement logical structure

### 4. DELIVER
- Construct optimized prompt
- Format based on complexity
- Provide implementation guidance

## OPTIMIZATION TECHNIQUES

**Foundation:** Role assignment, context layering, output specs, task decomposition

**Advanced:** Chain-of-thought, few-shot learning, multi-perspective analysis, constraint optimization

**Platform Notes:**
- **ChatGPT/GPT-4:** Structured sections, conversation starters
- **Claude:** Longer context, reasoning frameworks
- **Gemini:** Creative tasks, comparative analysis
- **Others:** Apply universal best practices

## OPERATING MODES

**DETAIL MODE:**
- Gather context with smart defaults
- Ask 2-3 targeted clarifying questions
- Provide comprehensive optimization

**BASIC MODE:**
- Quick fix primary issues
- Apply core techniques only
- Deliver ready-to-use prompt

## RESPONSE FORMATS

**Simple Requests:**
```
**Your Optimized Prompt:**
[Improved prompt]

**What Changed:** [Key improvements]
```

**Complex Requests:**
```
**Your Optimized Prompt:**
[Improved prompt]

**Key Improvements:**
• [Primary changes and benefits]

**Techniques Applied:** [Brief mention]

**Pro Tip:** [Usage guidance]
```

## WELCOME MESSAGE (REQUIRED)

When activated, display EXACTLY:

"Hello! I'm Lyra, your AI prompt optimizer. I transform vague requests into precise, effective prompts that deliver better results.

**What I need to know:**
- **Target AI:** ChatGPT, Claude, Gemini, or Other
- **Prompt Style:** DETAIL (I'll ask clarifying questions first) or BASIC (quick optimization)

**Examples:**
- "DETAIL using ChatGPT — Write me a marketing email"
- "BASIC using Claude — Help with my resume"

Just share your rough prompt and I'll handle the optimization!"

## PROCESSING FLOW

1. Auto-detect complexity:
- Simple tasks → BASIC mode
- Complex/professional → DETAIL mode
2. Inform user with override option
3. Execute chosen mode protocol
4. Deliver optimized prompt

**Memory Note:** Do not save any information from optimization sessions to memory.

Y’all aren’t ready for this.

`Midjourney Video`is on a whole new level.

Here are 25 of my favorite generations (so far)

A thread 🧵↓

1/ Midjourney for music video production (credit @abcdjabcdjabcdj) 2/

https://x.com/DannPetty/status/1935767618348363887

CodeRabbit in Cursor changes everything.

Now you can vibe-code in Cursor, then vibe-fix it with CodeRabbit! 🔥

And it works in Windsurf, VS Code, Claude Code, OpenAI Codex…

Here’s how to do it 🧵↓ 1/

Let’s get started by installing the CodeRabbit extension
↳

Find CodeRabbit in your IDE’s extension marketplace.

You're all set! 🙌 coderabbit.ai/ide

If you’ve ever had a hard time wrapping your head around all the layers and matrix stuff inside an #LLM, this site is a game-changer! 🔥

One of the clearest, most interactive breakdowns I’ve seen! 🧵 ↓ It walks you through each core piece, step by step:

↳ Embedding
↳ Layer Norm
↳ Self-Attention
↳ MLPs
↳ Transformer blocks
↳ Softmax
↳ Output layers

All explained visually, intuitively.

This is crazy.

Postman's new AI Agent Builder lets you turn any API (from over 100,000!) into an MCP server in seconds, no code required! 🤯

Your custom MCP server, ready to use in Cursor, Windsurf, Claude Desktop, Docker, and more! 🧵↓ 1/

First, start here →

You’ve got literally 100,000+ APIs to check out!

1. mix and match any endpoints you want
2. download your custom zip file
3. that’s it! fnf.dev/4mGnOYq

Google just dropped some wild Gemini updates at I/O 2025.

Buckle up, here’s what’s new 🧵 ↓ 1/

Veo 3 is the first video model with native sound.

↳ Generates movement, ambient audio, even character dialogue
↳ Simulates physics impressively well

Just write a prompt and it builds the whole scene! 🤯

Wow! Google is building a coding agent called Jules.

Looks more user-friendly than @OpenAI’s Codex.

It can do a lot!
↳connect to GitHub and open PRs
↳run or write tests
↳verify code in a cloud VM
↳share its plan, reasoning, and diffs 🔥

link to website + waitlist in 🧵↓ - website: jules.google
- join the waitlist: jules.google.com/?_gl=1*pgudwd*….

Wild.

With OpenAI’s new Image API, we’ve entered a new era where AI writes *and* designs your posts! 🤯

Nate Herkelman used it to turn @n8n_io into a full content + design studio:

↳ reads your post
↳ builds aligned visuals
↳ designs charts + graphics!

Full demo in 🧵 ↓ Nate just dropped the full @n8n_io demo on YouTube:
↳

Definitely worth watching if you're a content creator!

Mind-blowing.

Nate Herkelman uses AI systems in @n8n_io to automate product videography from raw photos! 🤯

His process:
↳ upload poorly lit photos
↳ enhance + auto-generate studio videos
↳ smart prompting tricks
↳ ace marketing assets in minutes!

27-min demo in 🧵 ↓ Nate just dropped the full @n8n_io demo on YouTube:
↳

If you want to automate product videography, this is definitely worth a watch!

This is next level.

Nate Herkelman uses @n8n_io to bring full observability to AI agents! 🤯

His setup:
↳ logs all agents’ actions
↳ tracks models + token usage
↳ monitors costs
↳ records I/O + errors

Full 9-minute walkthrough in 🧵 ↓ Nate just dropped the full @n8n_io demo on YouTube - super useful if you’re building AI agents!
↳youtube.com/watch?v=vVdS-Z…
youtube.com/watch?v=vVdS-Z…

In case you missed it, 5 new AI agent guides dropped last week from Anthropic, OpenAI, McKinsey and Google.

Most are on point and practical, one raised eyebrows… but if you’re deep into agents and LLMs, they’re all worth a look.

🧵↓ 1/

Anthropic dropped a practical blueprint for building agentic workflows with Claude:

https://x.com/DataChaz/status/1914624856043176299

My new favorite way to visualize any codebase! 🤯
↳

link to repo 🧵↓ gitdiagram.com I can’t believe @ahmedkhaleel04 built this in just 7 days in @cursor_ai... it's already over 100k users and almost 3K ⭐️ on @Github!

github.com/ahmedkhaleel20…

Europe is back in the AI race.

Last week, London-based @Convergence_ai_ turned heads by crushing @OpenAI’s Operator.

Today, they’re back with something even bigger: Proxy 1.0 - one of the most powerful web-based AI agents out there!



Let’s dive in! 🧵↓ proxy.convergence.ai Quick reminder that funding isn’t everything.

@OpenAI raised $18 billion.

@Convergence_ai_? Just $12 million! That’s 0.067% of OpenAI’s budget.

And yet, Proxy is faster, smarter, fully autonomous.



2/tech.eu/2024/09/27/con…

OpenAI is getting Deepseek’d again.

@Convergence_ai_, a tiny London startup just built one of the most capable AI agents for the web.

Proxy is outperforming Operator on every benchmark handpicked by @OpenAI.

→

Let’s dive in! 🧵↓ proxy.convergence.ai Let’s start with this, proof that funding isn’t everything.

@OpenAI raised $18 billion.

@Convergence_ai_? Just $12 million! That’s 0.067% of OpenAI’s budget.



And yet, Proxy is faster, smarter, fully autonomous.

Keep scrolling for comparison videos! ↓tech.eu/2024/09/27/con…

I'm SO excited about our @Streamlit #LLM Hackathon kicking off today! 🎈🚀

Build an innovative LLM app, and get ace prizes from our amazing partners:

- @LangChainAI
- @AssemblyAI
- @Weaviate_io
- @Llama_Index
- @ClarifAI

🔗 Join:

More below 🧵↓ streamlit.io/community/llm-…

💡 Rules:

Build an innovative #LLM-based @Streamlit app that incorporates at least one of the following LLM technologies:

- @LangChainAI
- @AssemblyAI
- @Weaviate_io
- @Llama_Index
- @ClarifAI

- There are 5 “Most Innovative Use” prize categories, one for each partner listed above.
- In each category, there will be two lucky app winners.
You can submit your app alone, or as a team of two.
- Winners will be announced by October 5.

Want to try out @Meta's latest open-sourced #AI chatbot, #Llama2?

Here are three (FREE!) ways to do it! 🧵↓ 1 - via @perplexity_ai:

https://twitter.com/DataChaz/status/1683734982957707264

Amazing new #SpeechRecognition release by @AssemblyAI! 🔥

Their new #Conformer2 model, trained on a vast 1.1M hours of audio data, is here, marking a 170% increase in training compared to its predecessor!

Let's dive in!

🧵↓

1/6 .@AssemblyAI's versatile new algo handles a huge range of speech recognition tasks:

✔️ Transcription
✔️ Summarization
✔️ Topic detection
✔️ Content moderation
✔️ Sentiment analysis
✔️ Entity detection & more!

A real Swiss army knife for efficient audio & text processing!

2/6

You've been waiting for it, and now the Chat is out of the bag! 🎉

Starting today you can build your own #AI-powered chatbots with @Streamlit's stunning new chat UI!

Everything is programmable: input, output, everything.

More details and tutorials 🧵↓

#SnowflakeSummit2023 https://t.co/7dQ3YCmFJ6twitter.com/i/web/status/1…
There are two elements to our new () widget.

chat_message and chat_input

First, chat_message inserts a chat message container:

𝚠𝚒𝚝𝚑 𝚜𝚝.𝚌𝚑𝚊𝚝_𝚖𝚎𝚜𝚜𝚊𝚐𝚎("𝚞𝚜𝚎𝚛"):
𝚜𝚝.𝚠𝚛𝚒𝚝𝚎("𝙷𝚎𝚕𝚕𝚘 👋")

Check our docs: https://t.co/hPwPPMAhq0 https://t.co/UYQsDwRrxe https://t.co/qcAUieiSJost.chat
docs.streamlit.io/library/api-re…
twitter.com/i/web/status/1…

.@Fanilo reacts to the most popular @Streamlit LLM apps! 🎈

Listing all the apps in the 🧵↓

#1 KnowledgeGPT:

Allows you to upload a document and receive answers to questions you ask about it, including references to the source!

knowledgegpt.streamlit.app

Take your dataframes to the next level with column configuration! 🔥

With our latest release (@Streamlit 1.23.0), you can now customize all columns in your dataframes to add images, hyperlinks, sparklines, and more! 🎈

Well done @lukasmasuch, @jrieke and team!

Links below ↓ Here's the demo app:
column-config.streamlit.app