Tweet

🇷🇴 cristi

Follow @CristiVlad25

Jan 9 • 8 tweets • 3 min read

How about some AI straight into your terminal?
Just say: gpt3 <your_command>.

(thread)

1. Go to my GitHub and clone this repository.

github.com/CristiVlad25/g…

2. Copy gpt3.sh to /root or another convenient location.

3. Make sure to edit line #4 to reflect your OPENAI API Key. Save. Then: chmod +x gpt3.sh

4. For convenience of access, add an alias to /.bashrc located in /root, such as: alias gpt3=/root/gpt3.sh

5. Use: gpt3 <Your prompt>

6. I've hardcoded some of the parameters (temperature, presence penalty, frequency penalty, max token, top_p) for convenience. In the future, if there is interest, I might develop this further.

@CristiVlad25

7. I'd appreciate hitting the "star" button on the repo at github.com/CristiVlad25/g…

For more tweets and threads: like, retweet, and follow me @CristiVlad25.

#gpt3 #chatgpt #ai #linux #machinelearning #productivity

• • •

Missing some Tweet in this thread? You can try to force a refresh

This Thread may be Removed Anytime!

Twitter may remove this content at anytime! Save it as PDF for later use!

More from @CristiVlad25

🇷🇴 cristi

@CristiVlad25

Jan 9

How to learn reverse engineering fast. A Practical approach.

(thread)

1. Go to crackmes.one and start playing with the challenges there.

2. Study from the book "Reverse Engineering for Beginners". You can do it for free at: beginners.re

Read 5 tweets

🇷🇴 cristi

@CristiVlad25

Jan 8

The most frequent vulns I found in 80+ pentests in 2022.

(thread)

1. Rate Limit Bypass

In more than half of the pentests I conducted, I found that there were no rate limits imposed on login functions or authentication mechanisms. This lack of rate limiting, combined with weak password policies, can lead to accounts being easily compromised.

2. Session Token not Invalidated upon Logout

This vuln is caused when the session token is not invalidated at user logout. If the validity of the token or the time to expiry is long, this can lead to the token being leaked, potentially resulting in a data breach or other vulns.

Read 5 tweets