Benjamin is a personal assistant that can help you build a web server in any language by only providing the template you want to use. Primary supported backend programming language: Rust
Create a .env
file in the root of the project and add the following:
OPEN_AI_ORG=<YOUR_ORG>
OPEN_AI_KEY=<YOUR_KEY>
GPT_MODEL=<YOUR_MODEL> # gpt-3.5-turbo, gpt-4
# I recommend using gpt-4 since it is the latest model and has more capabilities.
# Moreover, I have tested the assistant with gpt-3.5-turbo and the results were not quite satisfying due to
# many errors and bugs in the code generated by the assistant.
Moreover, add these to .env file for the assistant to locate the files and directories:
- The absolute path to the template you want to use.
CODE_EXECUTE_TEMPLATE_ABSOLUTE_PATH="/home/username/template/src/template.rs"
- The absolute path to the main file of the project you want to create.
EXEC_MAIN_ABSOLUTE_PATH="/home/username/template/src/main.rs"
- The absolute path to the api endpoints schema file.
API_SCHEMA_ABSOLUTE_PATH="/home/username/template/src/schemas/"
- The absolute path to the web server project.
WEB_SERVER_PROJECT_ABSOLUTE_PATH="/home/username/template/src/"
Note: You must create the directories and files before running the project.
To use the assistant, run the following command:
cargo build --release
./target/release/benjamin-assistant
or if you want to use the CLI, run the following command:
cargo run
Then you can start the conversation with the assistant by answering its prompts.
- The assistant has not been tested for building a frontend project yet.
- The prompts only allow the model to generate responses according to the given code template.