Evolving LLM-powered tool to assist pentesters and security researchers
For now imple assistant with RAG capability to utilize pdf as part of inference
Once RAG works reasonably well, should proceed to fine-tuning small model with sanitized pentest manuals
- [ollama|https://ollama.ai/]
- install dependencies:
poetry install - init/load vectordb:
poetry run python rag_vdb_load.py - startup ollama webserver:
ollama serve - run chainlit:
poetry run chainlit run pentestpal.py -w