Looking into adding functionality for allowing AgentGPT to output information to files

[Doc0x1]

Hey everyone, a friend of mine recently introduced me to this project and after playing around with it a bit, I realized it doesn't have any functionality yet for outputting things to files (e.g. if a user wants some code/project to be output so that he can open it up in VS code later on or something) so I thought I'd start looking into working on something for that.

While I wouldn't call myself an expert at TS, I've had a lot of experience with JS and Python (as well as Docker) and I've already forked the project and put some instructions into the setup.sh file checking for a $HOME/agent-output directory, and if not found, then the script creates the directory. Then it just runs a command to bind mount the directory to the container using Docker (I am aware the project also uses docker-compose and that you can run it locally as well, but considering I'm using Docker for convenience, I figured I'd just focus on that part of it for now).

Now, since I am new to projects involving AI and I'm not all that familiar with which parts of the project deal with interpreting user input (among other things), I am hoping someone might find some time to give me some pointers as to which areas I should focus my efforts on so that I don't have to go looking through all of the project files one by one.

The main thing I'm not entirely sure about is how I'm gonna get the project to be able to recognize which things it should put aside and output to files. I'm considering using the AgentTask model with a new Boolean field value to signify if it should take the value field (or the info field..? not really sure which yet) and output it to a file.

Anyway, if you have any questions for me, tips, advice, or whatever, feel free to post here or get in touch with me via the email listed in my GitHub profile.

Once again, very intriguing (and impressive) project, and I hope I might be able to contribute some useful things to it. Thank you.