Cursor Chat History FAQs

Question 1

How do I set up the Docker container?

Accepted Answer

First, run the extraction script (`cursor_history_extractor.py`) to create the LanceDB database. Then, build the Docker image and run the container, making sure to correctly map the database directory on your host machine to `/data/cursor_chat_history.lancedb` inside the container. Configure the `OLLAMA_HOST` environment variable to point to your Ollama instance.

Question 2

What is Cursor Chat History?

Accepted Answer

Cursor Chat History is a tool that lets you search and analyze your past conversations within the Cursor IDE. It extracts prompts, creates vector embeddings using a local LLM (Ollama), stores them in a LanceDB vector database, and provides a Dockerized FastAPI API for easy searching.

Question 3

What are the requirements to run Cursor Chat History?

Accepted Answer

You'll need Python 3.7+, Docker, Ollama installed and running, and access to your Cursor workspace storage directory. You also need to pull the `nomic-embed-text` model in Ollama. Python dependencies are listed in requirements.txt. Ensure your database directory has read/write permissions.

Question 4

What are the benefits of using this tool?

Accepted Answer

This tool enables you to reuse and analyze your past Cursor IDE conversations for Retrieval Augmented Generation (RAG), better understand your workflows, and quickly find previously generated code or ideas. It's also self-hosted and uses local LLMs for enhanced privacy and control.

Question 5

How does it work?

Accepted Answer

The tool extracts chat history from Cursor's local database, generates embeddings using Ollama's `nomic-embed-text` model, stores them in LanceDB, and then exposes a search API via a Dockerized FastAPI application. This allows you to perform similarity searches on your chat logs.

Cursor Chat History

概要

主な機能

ユースケース