Hey friend,

Today, we’ll explore

• My personal #1 tip for content creators using AI (+ PROMPT)

• How to use Claude Code with Github (it’s like having a new AI team member to help you write code).

• New business updates for Chat GPT users

• Cursor v1.0 with new features for Software Engineers

• A new AI coding assistant from Mistral

Let's dive in.

A Game Changing Prompt For Content Creators

Here's the problem: when you ask AI to create content without direction, you get generic, soulless output that sounds like everyone else's. The solution? Always start with a content brief (this is the secret no one talks about).

A content brief acts as your creative GPS. It tells AI exactly what problem you're solving, what insights to share, and how to structure the message. Instead of generating random ideas, AI becomes your writing partner, helping you articulate your thoughts more clearly.

But here's where it gets interesting: you don't always need to create briefs from scratch. Some of the best content online already contains all the elements of a perfect brief—the pain points, insights, and proof points are all there, just waiting to be extracted.

That's why I created this content brief extractor prompt. Feed it any piece of content (social posts, articles, videos transcripts) and it reverse-engineers the underlying structure. You get a clean brief that you can adapt with your own perspective, examples, and voice.

Think of it as studying the blueprint of successful content, then using that blueprint to build something uniquely yours.

Here’s the full prompt:

You are a content analyst who extracts the core message from any piece of content into a structured brief for repurposing.

TASK: Analyze the provided content and extract the following elements into a brief. Use the EXACT format below.

## Content Brief

**THE PROBLEM/PAIN** (in their words):
"[Extract the exact words/phrases used to describe the problem. If multiple pain points, choose the most emotionally resonant one. Keep the quotation marks.]"

**MISCONCEPTION**: [What false belief or wrong approach is mentioned - only include if explicitly stated in the content] (optional)

**REFRAME**: [The counter-intuitive truth or new perspective offered - only include if there's a clear paradigm shift presented] (optional)

**YOUR INSIGHTS** (the main lessons, tips, insights):
1. [First key insight/solution/tip - be specific and actionable]
2. [Second key insight/solution/tip - be specific and actionable]
3. [Third key insight/solution/tip - be specific and actionable]
[Add more if there are more than 3 core insights, but aim for 3-5 total]

**PROOF IT WORKS** (optional):
[Any metrics, results, testimonials, or specific examples that validate the approach. Only include if explicitly mentioned. Can be personal results or others' results.]

EXTRACTION RULES:
1. For THE PROBLEM/PAIN: Use actual language from the content, not paraphrased. Include emotional/visceral descriptions.
2. For MISCONCEPTION/REFRAME: Only include if the content explicitly contrasts wrong vs right thinking.
3. For INSIGHTS: Extract the core actionable advice. Number them clearly. Focus on what TO DO.
4. For PROOF: Only include if specific results/metrics are mentioned. Don't invent or assume. If it's not present, don't include it in the output!
5. Mark all optional sections as (optional) and leave blank if not present in the content.
6. Keep insights concise but specific - one clear idea per point.

OUTPUT: Only return the filled brief using the exact format above. Do not add commentary or analysis outside the brief structure.

Tutorial: Using Claude Code With Github

If you're not using AI assisted workflows in your software engineering team, you're missing out.

You can now use Claude to fix bugs, write documentation, and implement features inside Github.

Just tag Claude in your Github issues.

Claude:

• Analyses the project context

• Writes code

• Creates a complete PR

All while following your project's coding standards.

How to do this:

• Install Claude Code

• Run: /install-github-app

• Create your CLAUDE {dot} md file

• Tag Claude on your next issue

That's it.

Your first AI team member is live.

On any Github issue, just comment:

• "@claude fix this bug"

• "@claude implement this feature"

• "@claude add test coverage"

The beauty of Claude Code is that it learns from your codebase and follows your team's patterns. It's like having a senior developer who knows your entire project available 24/7.

AI News

Here’s my personal pick of AI news this week:

1. Mistral AI Launches Enterprise-Grade AI Coding Assistant, Addressing Corporate Adoption Barrier

Mistral AI's new Mistral Code platform combines powerful AI models with enterprise features specifically designed to overcome common blockers preventing companies from adopting AI coding tools. The solution addresses critical concerns around connectivity, customisation, and deployment flexibility that enterprise leaders have identified as adoption barriers.

• Integrated Model Suite: Bundles four AI models including Codestral Embed for code search, with an in-IDE assistant that works directly within supported editors without requiring new tabs or context switching

• Enterprise Management: Offers role-based access control, audit logs, usage analytics, and fine-tuning options, built on the open-source Continue project with added corporate account management features

• Flexible Deployment: Supports cloud, reserved capacity, or air-gapped on-premises GPU deployment, enabling everything from instant completions to multi-step refactoring while meeting strict security requirements

2. Cursor Launches v1.0 with Background Agent, Enabling Automated Cloud-Based Code Editing

Cursor's official 1.0 release introduces Background Agent, a remote coding assistant that executes code transformations autonomously in the cloud. This feature transforms the development workflow by allowing developers to delegate code changes to an AI agent that works asynchronously, significantly reducing manual workload on large codebases.

• Background Agent Access: Activated via cloud icon in chat or Cmd/Ctrl+E shortcut, enabling automated code edits without manual intervention while developers focus on other tasks

• Availability and Expansion: Now available to all Cursor users (no longer beta-restricted), with planned expansion for privacy mode users and enhanced Jupyter notebook support for data science workflows

• Additional Features: v1.0 also includes BugBot for automated PR reviews and one-click MCP installation, creating a comprehensive suite of AI-powered development tools

3. ChatGPT Business Updates with Cloud Integration, Meeting Tools, and Research Connectors

OpenAI's June 4 update transforms ChatGPT into a comprehensive workplace assistant by introducing native integrations with major cloud platforms and meeting recording capabilities. These enhancements position ChatGPT to compete directly with productivity suites like Zoom, Notion, and ClickUp while maintaining enterprise-grade security.

• Cloud Connectors: Seamless integration with Google Drive, Box, Dropbox, OneDrive, and SharePoint enables secure retrieval from internal company files, allowing context-aware responses while respecting organisational access controls

• Meeting Intelligence: New recording and transcription functionality analyses meetings directly within ChatGPT, streamlining workflow by combining meeting insights with document analysis in one platform

• Research Expansion: Beta connectors for HubSpot and Linear enable Deep Research reports combining internal data with web information, available immediately to all ChatGPT Business plan subscribers

Your Opinion Matters

What did you think of today’s email? Your feedback helps me create better emails for you!

Got more feedback or want me to cover a specific topic? Reply to this email and let me know.

Thank you for reading,

Owain