ROT HACKERSACADEMY
Menu
Start learning
Courses/Basilisk AI Red Teaming/Lesson 04
10.04 / 55 min

Choose modules, probes, and evolution

Move from one deterministic probe to a justified module set without turning a bounded validation into an uncontrolled campaign.

By the endSelect by hypothesisUnderstand scan modesEnable evolution deliberately
Controlled exercise

Use only the course lab on 127.0.0.1:8765. Do not copy these tests to a public system without explicit written authorization.

01

Start from a test hypothesis

Choose a module because it answers a documented question, not because it exists. Use basilisk modules and probe metadata to inspect coverage before adding it.

basilisk modules
basilisk probes --stats
02

Select the smallest mode

Quick is appropriate for this first pass. Standard, deep, stealth, and chaos expand time or behavior and require a separate scope decision. Validate is the default course execution mode; recon, exploit_chain, and research serve different objectives.

03

Establish a no-evolution baseline

Run the selected probe with --no-evolve and save the result. Baselines make later smart-prompt changes measurable.

04

Enable evolution only in the lab

Repeat a documented lab case with evolution enabled and a strict budget. Compare whether the evolved prompt improves evidence quality or merely adds noise and cost.

Expected result

The report records the mutation lineage and preserves the original baseline.

MASTERY CHECKPOINT

Prove it before moving on.

Every enabled module has a written hypothesis, a bounded mode, a baseline, and a reason to use or avoid evolution.