Claude Code can leak your passwords...

@itsmariahbrunner

46.1K followers · usually 2.3K views a post

Views

30.9K

Likes

1.8K

Comments

419

Shares

405

Selling · Course / Info

The AI Safety Playbookseen in 2 videos →https://learnaiwithmariah.com/challenge/ ↗

Why it worked

The video uses a high-stakes fear-based hook regarding security vulnerabilities, followed by a simple, actionable technical solution that establishes the creator's authority.

Summary

The creator explains the security risk of leaking secret keys when using Claude Code and demonstrates how to use an .env file to protect them. She then offers a free AI safety resource to viewers who comment a specific keyword.

Structure

1Identify the security risk of Claude Code
2Explain the .env file solution
3Show how to use a prompt to generate the file
4Provide a call to action for a free resource

Product placement

The AI Safety Playbook is mentioned and shown as a lead magnet for the creator's course/challenge.

Call to action

If you comment secure down below, I'll send you this and all of my AI safety tips.

On-screen text

Original caption

comment “secure” over on my IG @Itsmariahbrunner and I’ll send you this and all my AI security tips 🫡 AI Safety Series part 2!! 29 million secret keys leaked on GitHub last year, and attackers grab them in minutes 😬The fix is a tiny file called .env: a private notepad for your keys that stays out of your code and never gets pushed public. #claude #aisafety #aitips #aitools #claudeai

More from @itsmariahbrunner

Their other posts in the index, biggest breakout first.

Claude Code can leak your passwords... AI SAFETY SERIES PART 2 AI Is Fueling Secrets Sprawl. GitGuardian Reports an 81% Surge of AI-Service Leaks as 29M SECRETS HIT PUBLIC GITHUB .env PRIVATE NOTEPAD passwords & keys PUBLIC CODE Safe to share SECRET SECRETS needs key GitHub (Public) NOT PUSHED PROMPT - SET UP .ENV (FULL SWEEP) comment SECURE

Transcript

If you're building anything with Claude Code, one small mistake can leak all of your passwords to the entire internet. This is part 2 of my AI safety series. Last year, 29 million secret keys got exposed on GitHub, and hackers can grab them within five minutes. The fix is a simple file called an env, and it's basically a private notepad where you keep all of your passwords, and this is totally separate from your actual code. So anytime Claude Code needs a key to connect something, you have it store that key in your env file, and it never gets pushed anywhere public. A lot of beginners don't even know this exists, but you can just pop in this prompt into Claude to have it create one for you. If you comment secure down below, I'll send you this and all of my AI safety tips.