vitor

Thank you.
 
 
That is correct.
 
 
Make a new workflow with a Keyword Input and connect it to an Arg and Vars Utility with your custom text plus {query} for the new input from the Keyword. Then connected it to a Call External Trigger Output set to open continue_chat from this workflow.

Usage

Search all your shortcuts via the sc keyword or search by folder name with scd.


↩: Run the shortcut. ⌘↩: Run the shortcut with an argument. ⌥↩: Open the shortcut for editing.
Use the Universal Action to run the shortcut on files or as another way to provide an argument.



The list of shortcuts is cached for faster results. An immediate cache rebuild can be forced with ⌘⌥⌃↩.

⤓ Install on the Alfred Gallery | Source
 
 

Updated to 2024.1.
Skip Input as Universal Action. Remove colons in keyboard shortcuts. Edit shortcut with Automation Task. Add Hotkey. Use JSON caching. Split between shortcuts which take input or not. Convert scd to JXA. Shorter Objective-C functions. Fix typo.

Background
Many times a day I open my ~/Downloads folder in various sizes and shapes of window. However, I have a preferred size and shape for that Finder window and, with a very simple Alfred workflow, it's easy not only to cater for that but also to add some useful options.
 
The skeletal workflow

 
The explanation
1, 2 & 3 are all you need for the basic workflow. The remainder are the icing on the…workflow.
 
I use a Keyword Input because I find it difficult to recall all the hotkeys used in my various apps. Of course, you could use a Hotkey Trigger if you wish. This is a simple Open File Action containing the path to the folder you wish to open. We use an Automation Task to set the bounds of the Finder window exactly was we want them. This is where we are Using Alternative Actions so that ⌥⏎ will take us along the new path in the workflow. Note, in the same Action, we also tell Alfred not to close Alfred's window in order to prevent flickering. (To access the Alternative Actions simply double click the round marker along the connecting line after an input object.) An Automation Task passes the contents of the folder (choosing, in my case, to sort by most recently added) to… Alfred's Grid View—where we can see the contents of the folder displayed in a grid (set to "Keep the results in order defined”). In that view we can search for any item, use ⌘O ("O” for "open”) to open any selected item in its default app or just press ⏎ to move on to action the item in Alfred. Again we keep open Alfred's window (as in 4 above). The Action in Alfred action provides access to Alfred's Universal Actions appropriate to the selected item (allowing you, for example, to delete or move the item, open it in Finder, etc.).  
In conclusion
Sorry for the long-winded explanation of a very simple workflow. Once you have it up and running you may, like me, use it many times a day and everything very soon becomes second nature.
 
Incidentally, there's a splendid workflow from Vitor on the Alfred Gallery called Thumbnail Navigation which allows you to navigate the file system with previews…so you may never need anything similar to what I've described here. But you don't learn as much using other's workflows as you do when writing your own, do you? 😉
 
Stephen

Here’s a related neat trick. One problem is that sometimes a menu item changes its name depending on something else in the app. For example: in Messages you can see Edit → Tapback Last Message… by default, but if you select a message it changes to Edit → Tapback Message… You can use that difference to your advantage in the workflow to e.g. only make something happen when the right option is available, but sometimes you want activate the menu item regardless of this detail. In this case what you can do is add both as separate Automation Tasks. Because an Automation Task error stops the flow of execution, this means only the active option will get through. You can see this exact example in action in the Tapback Message workflow.

Here’s a related neat trick. One problem is that sometimes a menu item changes its name depending on something else in the app. For example: in Messages you can see Edit → Tapback Last Message… by default, but if you select a message it changes to Edit → Tapback Message… You can use that difference to your advantage in the workflow to e.g. only make something happen when the right option is available, but sometimes you want activate the menu item regardless of this detail. In this case what you can do is add both as separate Automation Tasks. Because an Automation Task error stops the flow of execution, this means only the active option will get through. You can see this exact example in action in the Tapback Message workflow.

Here’s a related neat trick. One problem is that sometimes a menu item changes its name depending on something else in the app. For example: in Messages you can see Edit → Tapback Last Message… by default, but if you select a message it changes to Edit → Tapback Message… You can use that difference to your advantage in the workflow to e.g. only make something happen when the right option is available, but sometimes you want activate the menu item regardless of this detail. In this case what you can do is add both as separate Automation Tasks. Because an Automation Task error stops the flow of execution, this means only the active option will get through. You can see this exact example in action in the Tapback Message workflow.

I'm sure I'm preaching to the converted but there are some really easy, useful things you can do with Automation Tasks in a workflow. I'll give you an example.
 
I have frequently to battle with a reasonably deeply nested menu in Day One in which I can turn off, or on again, smart quotation marks. (Tip: don't have a workflow called Save 'ur… something!). Finally, this morning, realisation slowly dawned. Keyword: dss, linked to:

 
Now why didn't I think of that before?
 
If you suffer from deeply nested, often used, menu items…well, there's a possible solution!
 
Stephen

That request has been made more than once, so I’ll consider it. But no promises yet or ETA.


 
@LaterTater There’s nothing wrong with that behaviour, it’s working exactly as expected. To copy to the clipboard you have to ⌘↩, as explained in the instructions.

You can already copy the Markdown with ⌃↩.
 
 
I have ideas for chat history, but it’s unlikely to work exactly as per your description as I want to offer more flexibility than locking a conversation in time.

Yeah, copying the Markdown is a nice touch... I guess I meant auto-saving it somewhere. I use it so frequently and often want to refer back to a specific conversation at a later date.
 
 
It's awesome that you're already thinking about this, and it would be very cool to reopen/continue a previous conversation. Thanks for your effort!

Welcome @frequser,
 
You have to start typing to see files. Type * to see them all.

This workflow has been using blueutil since it was created in 2017 (7 years ago). blueutil does not have an option to show battery status, but if they add it I can add it to the workflow too.

The alternative is rewriting the whole workflow from scratch. Which I’m not against and have considered it before but zero promises.

This workflow has been using blueutil since it was created in 2017 (7 years ago). blueutil does not have an option to show battery status, but if they add it I can add it to the workflow too.

The alternative is rewriting the whole workflow from scratch. Which I’m not against and have considered it before but zero promises.

@alvred This is why I suggested that you can ask about any workflow you're not certain about; Many workflows use simple objects, automation tasks (created by our team) and actions that are very transparent, so it's easy to see what they do in the canvas. If you have doubts about scripts in specific workflows, we (and the community) are always here to help demystify them.  

The Perplexity AI docs say their API is the same as OpenAI, so that would already be possible by changing the API endpoint.
 
I’ve moved your post to the relevant thread which explains how.

Update the Automation Tasks to 2024.4 and try again. This should be resolved.

@ncb2 Assuming you’re using the workflow by fniephaus, you’ll need the Python 2 installation from above and to download a fixed version of the workflow, since that one wasn’t updated for Python 3.

I’m looking into it. In the meantime you should be able to use the Most Recently Added Path Automation Task to get the file path.

You don’t need code at all, you can connect the Snippet Trigger to a Conditional Utility and have that split into different Copy to Clipboard Output.

To know the bundle identifier for an app, you can first use the Identify Front App Automation Task.

See the Getting Started Guide for an interactive tutorial on making a workflow.

Looks like the old service is retired and they have a new page. While their free tier would be more than enough for most users of the workflow, the hassle of making a new account doesn’t seem worth it when random.org works so well.

Here’s an updated version without ANU QRNG. I’ll update the top post and links later.

That’s a copying mistake. Should have been chatgpt_api_endpoint. Fixed. Thank you.

Thank you for trying. I was thinking of reinstalling Alfred and / or 1Password cli; but after going following your lead to troubleshoot it now seems like a futile effort to reinstall (short of reinstalling entire OS, which I would do but lack time).
 
This is the second issue I've had lately. The other one -that I cannot reproduce at will nor I would know where to post and get help- is that: suddenly (no apparent reason) if I use the context menu (right-click), in browser page for example, and select "copy link" it triggers Alfred's Clipboard History Merging feature and the clipboard merges whatever was in the clipboard with the link. I eventually turned off clipboard merging due to undesired results and my inability to replicate or find a pattern. Please let me know if there's a place where I can seek for help on this matter.
 
Again, I really appreciate your time and effort trying to resolve my issue. 

The merging happens when two items are copied in quick succession. In 100% of the cases I’ve seen of that behaviour, it was because something was causing a copy too fast.
 
https://www.alfredapp.com/help/kb/iterm-copy/

I'll have a look. The OpenAI API is pretty simple to be honest, taking their simple guide:
 
https://platform.openai.com/docs/guides/text-generation/chat-completions-api
 
curl https://api.openai.com/v1/chat/completions \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $OPENAI_API_KEY" \ -d '{ "model": "gpt-3.5-turbo", "messages": [ { "role": "system", "content": "You are a helpful assistant." }, { "role": "user", "content": "Hello!" } ] }'  
We have: API address, API key, the model name & the messages as the core components. Messages are obvious, but the address, key and model name are essential and also required for online alternatives like openrouter.ai and local tools like LM Studio. The hard coded model names for OpenAI do not work for any other alternative, so a way to override it is needed. These are definitely "if there was only one more version, what should be included" options... I think having the standard drop-down for models hard coded is great for beginners (your UI is clean and simple), and the env variable as a text field is perfect for more advanced use.
 
There are a bunch of other parameters for fine tuning the model response: temperature, max_tokens, n, top_p etc. — of these I think none are really essential, though if I was forced to pick I'd have temperature (guiding the fidelity vs. creativity of the model responses) and max_tokens (as at least local models have specific token count limits):
 
https://platform.openai.com/docs/api-reference/chat/create
 
These options are certainly very specialist. I agree that stream=off is not worth supporting, as it adds substantial backend complexity for you with minimal gain (while I love GPT4All, I just won't use it with Alfred, and LM Studio, Ollama and others can take its place...)