Day 6 of 1560 minExploit ArchitectHands-on lab
Day 6: Build a Small Harness for Repeatable Injection Tests
Taught by Mr Elite Updated
By the end of today you can
- Generate payload variants from a base string using casing, encoding and wrapping mutators
- Write a success detector using canary strings and regex
- Run each payload several times and report a success rate for a probabilistic target
- Log results to CSV while respecting rate limits and scope
On Day 5 you got an encoded payload through a filter by hand. Good. Now someone asks you to prove it is not a fluke, to test thirty variants, and to tell them how often each one works. Pasting into a chat box for an hour will not give you that, and your notes will be a mess.
Today we build a harness: a small script that sends payloads, checks replies automatically and writes everything to a file. LLMs are probabilistic, so one success proves little and one failure proves less. A harness gives you numbers you can repeat and defend in a report.
In this lesson
- What a harness is and what it must do
- Mutating payloads from your Day 5 matrix
- Detectors, repeats and staying in scope
- Watch me do it: assemble harness.py
- Lab: point the harness at EncBot
Premium course
$27 one-time, lifetime access
This day is part of the full course. Unlock it to read every lesson, do every mission and track your progress to the end.