Problem
Mentioning the review actor in a top-level PR comment dispatches a review. Mentioning it in a reply inside an inline review thread does nothing, with no feedback.
GitHub sends thread replies as pull_request_review_comment events, not issue_comment. The GitHub webhook has no handler for that event, so it returns 200 and drops it. Replying in the thread the reviewer opened is the natural way to ask for a re-check, so this is the gesture most authors will try first.
Proposal
Add an on_pull_request_review_comment_created handler that publishes the same pr.commented event as on_issue_comment_created. The payload carries the same sender, comment.author_association, and comment.body fields, so the existing writer/mention/repo gates apply unchanged.
Also document that the operator app's webhook needs the "Pull request review comments" event subscription next to "Issue comments".
Problem
Mentioning the review actor in a top-level PR comment dispatches a review. Mentioning it in a reply inside an inline review thread does nothing, with no feedback.
GitHub sends thread replies as
pull_request_review_commentevents, notissue_comment. The GitHub webhook has no handler for that event, so it returns 200 and drops it. Replying in the thread the reviewer opened is the natural way to ask for a re-check, so this is the gesture most authors will try first.Proposal
Add an
on_pull_request_review_comment_createdhandler that publishes the samepr.commentedevent ason_issue_comment_created. The payload carries the samesender,comment.author_association, andcomment.bodyfields, so the existing writer/mention/repo gates apply unchanged.Also document that the operator app's webhook needs the "Pull request review comments" event subscription next to "Issue comments".