abhagsain/ai-cli

Get answers for CLI commands from ChatGPT right from your terminal

See my projects on my portfolio

💻 Currently Working at - Treebo.

Estimated Numbers of Hours Spent Coding (Doesn't include the time spent on debugging 😛)

Connect with me:

Languages and Tools:

ai-cli's Issues

Changing the default model.

Hello,

Thank you for your awesome tool.

Currently, this app uses the text-davinci-002 model by default. I was wondering if it is possible to add a custom config file/flag so we can change this to other models (e.g., code-davinci-002)

Thanks.

[Feature] Change OPENAI_API_BASE to a custom URL

To be able to use locally hosted LLMs.

Long answers are chomped

Firstly - bigup for this tool. its incredible how useful and wide-reaching it is. platform/ language/cli/iam roles ... everything. 🙏

But long answers are chomped:

I guess initially you might've not thought about such long responses, but now it seems its a killer tool that can answer a lot more!

Default model text-davinci-003 deprecated

Error: {"message":"The model `text-davinci-003` has been deprecated, learn more here: https://platform.openai.com/docs/deprecations","type":"invalid_request_error","param":null,"code":"model_not_found"}

The new models should be added to the ai model menu.

Error: Request failed with status code 400

Error upon ai auth

I get the following error

`/usr/local/lib/node_modules/@abhagsain/ai-cli/node_modules/@oclif/core/lib/command.js:41
delete this.globalFlags?.json;
^

SyntaxError: Unexpected token '.'
at wrapSafe (internal/modules/cjs/loader.js:915:16)
at Module._compile (internal/modules/cjs/loader.js:963:27)
at Object.Module._extensions..js (internal/modules/cjs/loader.js:1027:10)
at Module.load (internal/modules/cjs/loader.js:863:32)
at Function.Module._load (internal/modules/cjs/loader.js:708:14)
at Module.require (internal/modules/cjs/loader.js:887:19)
at require (internal/modules/cjs/helpers.js:74:18)
at Object. (/usr/local/lib/node_modules/@abhagsain/ai-cli/node_modules/@oclif/core/lib/index.js:6:19)
at Module._compile (internal/modules/cjs/loader.js:999:30)
at Object.Module._extensions..js (internal/modules/cjs/loader.js:1027:10)
`

Most replies are "Sorry can't answer that"

Basically 3 out of 4 attempts I do I receive as "sorry I can't answer that". I wonder how an answer like that is determined, when chatgpt3.5 itself provides it (for free especially)

Is it somehow related to a system prompt perhaps?

oclif 'update' fails

Steps

$ ai update --available

Expected

list of available updates via oclif/plugin-update

Actual

$ ai update --available
    Error: no s3 host is set

Regression

Not sure how this mechanism works speficifcally, e.g., if the oclif implementer needs to do some magic, or if it is an end-user thing. Therefore, I wasn't able to regress / debug much.

[Feature Request] Support for windows powershell

It seems like the commands delivered currently are linux (bash) only. It'd be cool to have the option for the model to deliver powershell commands when asking on a terminal running powershell.

Error: Request failed with status code 429

Don't know how to report it, so here is what I do and what I get:

❯ ai ask "unzip file"
... !
    Error: Request failed with status code 429

running ai auth creates a folder named "~" at my cwd

Steps to reproduce:

On an M1 Macbook pro running macos 13:
on a zsh terminal running oh-my-zsh:
run "ai auth"
paste the api key and then run ls on the zsh interface.

You will see a folder that has been created named "~" with the contents: ai/ and configstore/
Expected behaviour would have been that these filed were created at the home folder that is colloquially referred to as "~" in unix environments.

Error: Request failed with status code 429

When I enter the first command ai ask "Check process running on port xxx", it show the error message Error: Request failed with status code 429. My account are currently on the free trial.

I have no idea how to solve this issue. Any help will be appreciate.

GPT-4 ?

does this library support gpt 4 ?