[FEATURE] Add support for LLM fuzzing #75
Description
LLM Fuzzing
- Parameters right now are randomly chosen when there are no objects corresponding to the input in the objects bucket. Use an LLM to generate parameters than can be used for a target query / mutation
- As input, the LLM should understand the schema and possibly information regarding seen objects so that it can determine what is a possible input parameter
- Use Langchain to allow use of various LLM models