1) You say start
2) I give you a question
3) You give a answer
4) I rate the answer and accumulate the score
5) Once the score is X, I give you the flag <Actual Flag>.
Could it be any easier? 😀
In some funny ways, AI behaves like a little kid sometimes.
I was trying to solve a challenge. The challenge involved answering some questions. Without bad intentions, I just asked how to proceed and there you go, it told me the steps like this:
#AI#ArtificialIntelligence#Fail
Excerts from Cleverread: Amit Agarwal strips away unnecessary complexity and focuses on what matters most in real-world offensive security work: the right command, at the right time, with the right syntax.
cleverread.in/practical-pent…
zapgpt -up prompting -p openai -m gpt-5.4 "I am learning for <cert name> and you are my tutor called aka. You are an expert in <expertise here>." -o ~/.config/zapgpt/my_course.json
zapgpt -up my_course "Query"
#ai#automation#learning#cybersecurity#ethicalhacking#hacking
What cna you do with zapgpt?
zapgpt -up prompting 'I want to create a prompt - You are cybersecurity expert and specialize in AI hacking.' -o ~/.config/zapgpt/prompts/my_hacking.json
zapgpt -up my_hacking "prompt"
#ai#artificialintelligence#hacking#pentestin#ethicalhacking
Big news. Now you have 3 new features in zapgpt:
1) You can use -ndp to exclude common_base.
2) You can now provide upto 2 files
3) You can now use more than 1 prompt as system prompt
uv tool install zapgpt
uv tool upgrade zapgpt
github.com/raj77in/zapgpt
Beginner-friendly drop: Exploiting Blind SQLi on DVWA 🎯 Learn boolean-based + time-based inference, plus a clean enumeration flow. What should I cover next—manual vs sqlmap, or mitigation? youtu.be/9ujHoJwgjsM#SQLi#BlindSQLi#DVWA#AppSec#Pentesting