adding improvements to ai page docs

Author: commit111

docs/concepts/debug.md

@@ -8,8 +8,8 @@ sidebar_position: 650
 
 Defang includes an AI-driven tool in the CLI (command-line interface) to help you debug your cloud applications. The AI agent will use your service logs as well as the files in your project to help you identify and resolve issues.
 
-:::info
-We plan to expand the capabilities of the AI agent in the future. 
+:::tip
+Defang has another AI-driven tool called [`generate`](/docs/concepts/generate).
 :::
 
 ## How It Works
@@ -28,7 +28,6 @@ Here is a typical workflow in the [Defang CLI](/docs/getting-started) that will
 The AI debugger will not change your files. Instead, it will show you a suggestion, and it is up to you if you want to use it in your code.
 :::
 
-
 The AI debugger only kicks in when any service in a project fails to deploy. This could be because of a build failure, healthchecks failing, or a variety of other issues. 
 
 :::info

docs/concepts/generate.md

@@ -8,12 +8,12 @@ sidebar_position: 100
 
 Defang includes an AI-driven agent that translates natural language prompts to an outline for your project that you can then refine. The AI agent is available through the [Defang CLI](../getting-started#install-the-defang-cli.md).
 
-:::info
-The AI agent is currently in preview and only supports a limited set of prompts, varying in complexity. We plan to expand the capabilities of the AI agent in the future.
-:::
-
 We are working on expanding the range of supported prompts and improving the accuracy of the AI agent. If you have any feedback or suggestions, please let us know by [opening an issue](https://github.com/DefangLabs/defang/issues/new).
 
+:::tip
+This page is about Defang's AI-driven `generate` tool. We also have another AI tool for [debugging](/docs/concepts/debug). Don't get confused between the two!
+:::
+
 ## Example Prompts
 
 Here are some example prompts that the AI agent can help you with:
@@ -46,4 +46,4 @@ An HTTP GET endpoint that returns an HTML with a random meme from the imgflip AP
 
 ```
 A service that uses bullmq and redis. The compose file should also include a redis service. The code should set up a queue. It should also expose three endpoints with express: one to add a job to the queue, another to check how many tasks are in the queue, and a last one that runs on / which is just a healthcheck. The code should also include a worker that processes the queue by logging what's in each task. Use environment variables to configure the redis connection. No password.
-```
+```