Skip to content

feature/expand-review-scope: Adding review scope parameter - #38

Closed
jroachgolf84 wants to merge 2 commits into
mainfrom
feature/expand-review-scope
Closed

jroachgolf84 wants to merge 2 commits into
mainfrom
feature/expand-review-scope

Conversation

@jroachgolf84

@jroachgolf84 jroachgolf84 commented Oct 28, 2026 •

Copy link
Copy Markdown
Collaborator

Description

This PR adds a review-scope parameter that allows for a user to determine what Otto should consider. There are four options added as part of this PR:

  • entire-repo: default value, evaluates diffs across entire repo
  • airflow: just diffs in Airflow project
  • data-engineer: diffs in Airflow project + things like SQL files, etc.
  • platform-engineer: Terraform, IaC, etc.

Related Issue(s)

No related issues.

Breaking Change?

No, this is not a breaking change.

Checklist

  • I updated README.md and the action.yaml input/output descriptions if the action's interface changed
  • Lint YAML passes
  • I tested the change via the e2e/ harness (or explained why that wasn't possible)

@jroachgolf84
jroachgolf84 requested a review from a team as a code owner October 28, 2026 19:50
@github-actions

github-actions Bot commented Oct 28, 2026 •

Copy link
Copy Markdown

Otto Review

Adds review-scope, include-paths and exclude-paths inputs. On both GitHub and GitLab, these filter the PR/MR diff down to role-specific files before Otto reviews it.

1 major issue found

The scoping pipeline is well put together. Filtering happens before the line cap, the inline-comment filter runs against the scoped diff, the empty-diff check now runs before scoping, and the approve downgrade on partial reviews gives a mechanical backstop. Two earlier threads are still open and not fixed in this diff. First, apply-scope.sh is still called inside an if [[ "$(...)" == ... ]] test in both gather scripts, so a mistyped review-scope quietly falls back to a full, unscoped review. Second, a skipped run still leaves every action output empty. The new problem is in the skip path: it exits before post-review.sh runs, so Otto never reconciles any review it posted on an earlier push. An earlier REQUEST_CHANGES review can keep blocking the merge even though Otto no longer has anything to review.

Comment thread github/gather-context.sh Resolved
Comment thread action.yaml Resolved
Comment thread github/gather-context.sh Resolved
@jroachgolf84

Copy link
Copy Markdown
Collaborator Author

@milton-li - do you think it's going to make sense to update the --persona functionality to match this sort of behavior?

@jroachgolf84

Copy link
Copy Markdown
Collaborator Author

Closing this in favor of: #39 (review)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant