"It's like AutoGPT got a brew install
", made possible by Kurtosis.
Assuming you have Kurtosis installed, first start AutoGPT (replacing YOUR_API_KEY_HERE
with your API key):
kurtosis run github.com/kurtosis-tech/autogpt-package --enclave autogpt '{"OPENAI_API_KEY": "YOUR_API_KEY_HERE"}'
Then start interacting with AutoGPT:
kurtosis service shell autogpt autogpt --exec "python -m autogpt"
If kurtosis service shell autogpt autogpt --exec "python -m autogpt"
breaks for you then you might be on an older version of Kurtosis. Please use instead:
( echo "python -m autogpt" && cat ) | kurtosis service shell autogpt autogpt
We use the Redis
memory backend by default.
Follow along the official guide here
To pass any other configuration listed here; pass the argument like you pass the OPENAI_API_KEY
kurtosis run github.com/kurtosis-tech/autogpt-package --enclave autogpt '{"OPENAI_API_KEY": "YOUR_API_KEY_HERE", "RESTRICT_TO_WORKSPACE": "False"}'
Note - This package spins up AutoGPT using the Redis
backend by default. To use the local backend instead set MEMORY_BACKEND
to local
in args
. For pinecone
and milvus
you will need to get API keys for it and pass it. Weaviate
is supported both locally & remotely.
To run with an instance of Weaviate inside Docker run this using
kurtosis run github.com/kurtosis-tech/autogpt-package --enclave autogpt '{"OPENAI_API_KEY": "YOUR_API_KEY_HERE", "MEMORY_BACKEND": "weaviate"}'
Kurtosis supports the ALLOWLISTED_PLUGINS
configuration flag that AutoGPT
ships with. For example, to run the twitter
plugin do the following:
kurtosis run github.com/kurtosis-tech/autogpt-package --enclave autogpt '{"OPENAI_API_KEY": "YOUR_API_KEY_HERE", "ALLOWLISTED_PLUGINS": "twitter"}'
Under the hood, Kurtosis will download and install the package for you! As of now the following plugins are supported:
To add support for more plugins simply create an issue or create a PR adding an entry to plugins.star
.
NOTE - If you run AutoGPT with plugins; we are using our own image published at h4ck3rk3y/autogpt
. We are tracking the update to use the official 0.3.0 image in: kurtosis-tech#27.
Kurtosis has an extension available on VSCode that allows you to develop Kurtosis Starlark more efficiently. While develeoping this package locally run using -
kurtosis run . --enclave autogpt '{"OPENAI_API_KEY": "YOUR_API_KEY_HERE", "MEMORY_BACKEND": "weaviate"}'
This would upload the local package and run it instead of pulling it from GitHub.
Let us know in our Discord or on Twitter @KurtosisTech!
Feel free to create an issue on GitHub if you have any bugs or feature requets.