feat(judge_4): screenshot list in PR body + file-trigger gating (Slice C)#264
Merged
Merged
Conversation
…e C) Two small additions on top of slice B (#263): 1) PR review body lists screenshots produced by browser-judges. ``get_peer_review_status`` already aggregates the ``screenshots`` field across the panel (slice B); the panel-review composer now appends a collapsible ``<details>`` block listing each screenshot path + caption. Inline image rendering via GitHub-side hosting is a follow-up — would need to commit the files to a branch or upload via the asset API. Listing the paths still surfaces visual findings in the PR comment thread, and lets the operator open them locally with ``open .prompts/screenshots/<task>/<judge>/<x>.png``. 2) File-trigger gating on dev-server bring-up. ``_maybe_start_dev_server`` now also checks that the browser-judge actually has at least one frontend file routed to it. Backend-only PRs (judge_4 scope empty / all .py / all .md) skip the bring-up entirely — saves the ``pnpm dev`` startup + teardown on every backend PR. Frontend suffixes covered: tsx/jsx/ts/js/css/scss/ sass/html/svelte/vue/astro. Conservative when ``scoped_files`` is None (file routing didn't run): bring the server up. Cost-of-skip (visual regression slips) > cost-of-spin-up (wasted seconds). 7 new tests in test_judge_panel_browser_gating.py pin the gate's contract across the suffix set, the backend-only case, and the unknown-scope conservative default. 715/715 tests pass. Inline image rendering via GitHub asset upload is the remaining piece for full visual review in the PR thread.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Two additions on top of Slice B (#263): (1) panel summary now lists screenshots produced by browser-judges in a collapsible details block; (2) dev-server bring-up skips when the browser-judge has no frontend files in scope (saves pnpm dev startup on every backend PR). 715/715 tests pass.