Hi everyone!
Do you want to know who I talk to most at work? That’s right, robots.
We all know that smart workflows are incredibly useful for handling routine tasks, drafting content, and saving time. But what if you want to refine the output? Maybe you’d like to change the format, ask a follow-up question, or clarify how your guidance was interpreted? Maybe you just want to talk about what you’ve co-created without leaving Asana.
Well, you can! It’s pretty easy, too. You just need to add a single AI-powered rule to allow a back-and-forth with Asana AI, just like you would with a colleague.
Let me introduce you to AwesomeBot, a rule that you can have a back-and-forth conversation with, right where the work is happening.
OK, AwesomeBot is just a silly name for my AI conversation rule - you can name yours anything you like. In fact, it’s a terrible name, but that’s by design - it’s easy to remember, and makes it difficult (or even impossible) to accidentally loop AwesomeBot into the conversation when it’s not needed.
Now let’s look at how you can set this rule up.
The rule setup
Add this single AI-powered rule to any project to start the conversation:
When comment is added > Check if the comment begins with the term AwesomeBot > Add comment in task (Use AI)
That’s it! It doesn’t need to be built into a larger workflow, but it can.
The prompt
Here’s the guidance for AI that I use to make sure that AwesomeBot joins the conversation when it should, and butts out when it shouldn’t. Feel free to copy this guidance for your rule, and edit it as you see fit:
“Your job is to reply to the user only when the comment begins with the term AwesomeBot. If the comment does not begin with the term AwesomeBot, the comment is not directed at you, so you do not reply.
Examples of when you should reply:
- AwesomeBot, please summarize this comment thread.
- AwesomeBot, translate this to German.
- AwesomeBot, make the feature names bold.
Examples of when you should not reply:
- Summarize this.
- Translate this to German.
- We should make all the feature names bold.
- I like chatting with AwesomeBot, we are best friends
”
How it works
The mechanics of the rule work like this:
When a comment is added to a task in the project, the rule will check whether the comment begins with AwesomeBot.
If the comment is for someone else, the rule doesn’t run, so we avoid AwesomeBot jumping into the conversation when it’s not needed.
If the comment begins with AwesomeBot, in effect directly addressing the “bot”, the rule will run - and AwesomeBot will be allowed to reply.
Here’s a real-life example:
Why this is so useful
Think of this conversation rule as a hyper-flexible way to refine the output of your workflow, iterate on drafts or summaries, and adjust formatting on the fly. You can also ask follow-up questions directly and clarify how the AI has interpreted your instructions.
So the next time you’re thinking “I wish I had a friendly robot to chat with about the formatting of my email draft”, look no further than AwesomeBot - or another less ridiculously-named automation.
Give it a try, and start the conversation.
