What is zetsu?

Zetsu is a personal, offline tool for people who do offensive security work, like penetration testers or bug bounty hunters. The idea is simple: over time you build up a pile of your own notes, writeups, lab solutions, and saved blog posts about hacking techniques. Instead of digging through folders trying to remember which file had the right command, you can just ask a question in plain language and get an answer pulled from your own material. When you add your notes, Zetsu reads through them and pulls out the useful pieces, things like what was found, what action was taken, why, and what happened as a result. This means when you later search for something, you get a focused answer instead of a random chunk of text. It also combines two different search methods together, one that matches exact words like tool names or command flags, and one that matches based on meaning, so it can find relevant notes even if you do not phrase the question exactly like your original notes. You can talk to it in a few ways: a terminal interface for quick lookups during an actual engagement, a browser interface for slower study sessions, and a command line mode for scripting. It can pull answers using different AI providers, including Anthropic, OpenAI-compatible services like DeepSeek, or a locally run model through Ollama, so nothing has to leave your machine if you do not want it to. Setup involves cloning the code, installing Python dependencies, and adding an API key for whichever AI service you choose. You then drop your markdown notes into a folder, tell it where to find them in a config file, and run an ingest command before you can start asking questions. The project includes a benchmark the author ran across 910 security questions, reporting that it answered all of them and included working commands in the large majority of cases. It is aimed squarely at people who already have a personal knowledge base and want a fast way to query it, not a general purpose chat assistant.