## Background

- Pre-requisites
- Regex pattern fuzzing (see below)
- Regex parser comes first
- Then fuzzer later
- First steps
- Basic tree-walker algorithm of schemas (Metaschema modules)
- Second steps
- Metaschema and "Metaschema's schema"
- Building a fuzzer that fuzzes the Metaschema schema
- Find a way to encode model representations
- Perhaps bias different structures over others
- Equivalence classes
- Any field or flag with a string type with no additional string is just a generic string, can test it as such

## Goals

- Start simple, scale up
- Black-box first, white box earlier/later if we can

## Level of Effort
- The hard: pre-requisite or separate
- Verdict: keep it separate
- Fuzzing the data type patterns of different fields and flags
- The hard parts
- Black-box approach
- Scenario definition for the different scenarios
- The harder parts
- "Whitebox approach" -> see the implementation and look at it
- Actually reversing the implementation