[BUG] Generator attempts to submit anonymous metrics even after fatal generation failure

[sahillllllllll-bit]

Describe the bug.

The AsyncAPI Generator attempts to submit anonymous usage metrics even after generation has already failed with a fatal error. This results in an additional runtime error being printed, which is unrelated to the actual cause of the failure and pollutes the output.

Metrics submission should only occur after a successful generation or a clean shutdown, not after a hard failure.

Expected behavior

• Generation should abort immediately after a fatal error.
• No attempt should be made to submit anonymous metrics.
• No secondary or unrelated errors should be printed.

Screenshots

How to Reproduce

1. Create a non-empty output directory:

  mkdir out
  echo test > out/file.txt

2. Use the following AsyncAPI document:

asyncapi: '2.6.0'
info:
  title: Metrics Failure Test
  version: '1.0.0'
channels: {}

3. Run the generator:

asyncapi generate asyncapi.yaml @asyncapi/html-template --output ./out

4. Current Behavior
   The generator fails as expected due to the non-empty output directory, but then prints an additional error related to metrics submission:

Generator Error: "./out" is not an empty directory. You might override your work.
Skipping submitting anonymous metrics due to the following error: TypeError: fetch failed

🖥️ Device Information [optional]

• Operating System (OS):Windows(WSL)
• Browser:
• Browser Version:

👀 Have you checked for similar open issues?

• I checked and didn't find similar issue

🏢 Have you read the Contributing Guidelines?

• I have read the Contributing Guidelines

Are you willing to work on this issue ?

Yes I am willing to submit a PR!

[dosubot]

This is a known issue: the generator tries to submit anonymous metrics even after a fatal error, which leads to extra error messages that clutter the output. The main Generator class throws on fatal errors (like a non-empty output directory) and doesn't handle metrics submission itself—this logic is handled elsewhere, likely in the CLI entrypoint or a wrapper. There is currently no guard to prevent metrics submission after a hard failure, so the submission is attempted regardless of success or failure, which matches your observation.

To fix this, you'll want to ensure metrics submission is only triggered after a successful generation or clean shutdown. The right place to add this guard is in the CLI entrypoint or wherever the generator is orchestrated: set a flag or use error handling to skip metrics submission if a fatal error occurs. For reference, this behavior and its impact are discussed in this GitHub issue and confirmed in this Slack thread.

Let me know if you want guidance on where to add this check in the CLI code!

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}

   

[sahillllllllll-bit]

@Shurtu-gal @AayushSaini101 @Souvikns waiting for approval to submit pr:)

[1164148-7-eng]

/attempt #2010

I've claimed this issue. Working on it now! 🚀