Skip to content

chore: bump review image to 0.3.0 for structured diff view#22

Merged
fatherlinux merged 3 commits into
masterfrom
chore/bump-review-image-0.3.0
Jul 3, 2026
Merged

chore: bump review image to 0.3.0 for structured diff view#22
fatherlinux merged 3 commits into
masterfrom
chore/bump-review-image-0.3.0

Conversation

@fatherlinux

Copy link
Copy Markdown
Member

One-line bump of the reusable review workflow's default image from gatehouse:0.2.0 to gatehouse:0.3.0, so PR gates run the structured BEFORE/AFTER diff pipeline (#19, merged in #20). The 0.3.0 quay tag was published from git tag v0.3.0.

Kept separate from any code it gates, per the two-PR rule.

🤖 Generated with Claude Code

The reusable review workflow now runs the v0.3.0 pipeline, which feeds
agents a structured BEFORE/AFTER diff view instead of a raw unified
diff (#19, PR #20).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
@gemini-code-assist

Copy link
Copy Markdown

Note

Gemini is unable to generate a review for this pull request due to the file types involved not being currently supported.

@github-actions github-actions Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Gatehouse found 2 issues (2 high)

type: string
required: false
default: "quay.io/crunchtools/gatehouse:0.2.0"
default: "quay.io/crunchtools/gatehouse:0.3.0"

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

HIGH (Test Coverage): The workflow now uses Gatehouse image 0.3.0, which, according to the commit message, introduces a 'structured BEFORE/AFTER diff view' instead of a raw unified diff. This represents a significant change in core functionality (diff parsing and structuring, likely in diffview.py). There is no evidence in the provided diff that corresponding unit or integration tests have been added or updated to cover this new, complex logic, including various diff formats and edge cases.

Suggestion: Ensure comprehensive unit tests are implemented for the new 'structured BEFORE/AFTER diff view' functionality. These tests should cover: correct parsing of various unified diff formats into the structured view, handling of edge cases (e.g., empty files, only additions/deletions, renames, binary files, large diffs), and verification that the structured output correctly represents changes for agents. If diffview.py was modified, existing tests should be updated to reflect the new behavior.

Evidence:

-        default: "quay.io/crunchtools/gatehouse:0.2.0"
+        default: "quay.io/crunchtools/gatehouse:0.3.0"

type: string
required: false
default: "quay.io/crunchtools/gatehouse:0.2.0"
default: "quay.io/crunchtools/gatehouse:0.3.0"

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

HIGH (Documentation): The default image tag for the Gatehouse review workflow has been updated to 0.3.0, which introduces a significant behavioral change: agents now receive a structured BEFORE/AFTER diff view instead of a raw unified diff. This change in core functionality likely invalidates existing user-facing documentation (e.g., README, usage guides) that describes how Gatehouse processes diffs.

Suggestion: Update the project's main documentation (e.g., README.md, USAGE.md) to reflect that Gatehouse now provides a structured BEFORE/AFTER diff view to agents, rather than a raw unified diff. Explain the benefits or implications of this new structured view for users.

Evidence:

-        default: "quay.io/crunchtools/gatehouse:0.2.0"
+        default: "quay.io/crunchtools/gatehouse:0.3.0"

fatherlinux and others added 2 commits July 2, 2026 23:47
Addresses the review finding on this PR: the README never mentioned
that agents receive a structured diff view as of v0.3.0.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

@github-actions github-actions Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Gatehouse found 1 issues (1 high)

Comment thread README.md

All 8 agents run concurrently. Findings below 80% confidence are filtered out.

## How Agents See Changes

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

HIGH (Test Coverage): The provided diff only updates the workflow version and documentation for a new "structured BEFORE/AFTER diff view" feature (v0.3.0). The actual implementation of this feature, which involves complex diff parsing and structuring logic (e.g., handling unparseable diffs), is not included in the diff. This prevents assessing its test coverage for new or modified functions, complex logic branches (e.g., parsing failures, edge cases), and changed behavior.

Suggestion: Provide the code changes for the "structured BEFORE/AFTER diff view" feature (likely in diffview.py) to allow for a comprehensive test coverage review. Specifically, tests should cover successful parsing of various diff structures, edge cases for hunk parsing, and the described fallback to raw unified format for unparseable diffs.

Evidence:

Since v0.3.0, agents do not receive raw unified diffs. Each hunk is rendered
as a structured view: a BEFORE block (the old code, removed lines marked
`[-]`) and an AFTER block (the new code with real file line numbers, added
lines marked `[+]`), plus instructions to judge the direction of a change.
Protections *added* by a change are treated as fixes, not findings;
protections *removed* by a change are flagged. Diffs that cannot be parsed
fall back to the raw unified format.

@fatherlinux fatherlinux merged commit 0f85ef4 into master Jul 3, 2026
14 checks passed
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