Repository navigation
Verify generated pages with real Chromium rendering and pixel comparisons - #102
Merged
Conversation
owjs3901
marked this pull request as ready for review
October 19, 2026 15:32
owjs3901
merged commit Oct 19, 2026
fb0b798
into
devfive/3099154426bffbeef8b62999b80bb038/devfive-relay-c70423/devup-mcp-completeness
9 checks passed
owjs3901
deleted the
devfive/3099154426bffbeef8b62999b80bb038/real-render-verification
branch
October 19, 2026 15:32
owjs3901
pushed a commit
that referenced
this pull request
Oct 19, 2026
#101) * Require complete runtime evidence and catch responsive reading-order and clipping failures * Handle unavailable Windows symlink privileges in audit validation * Guard screen-plan frame inventories and bind navigation labels * fix: guard Figma section and asset coverage * Publish generated Figma scripts atomically during concurrent validation * Block unfinished visual comparisons and complete runtime audit reporting * Record final generated-script validation evidence * Record final rendered completeness audit validation * fix: reset button appearance and preserve hero image sources * feat: show devup-mcp version in bridge plugin and harden runtime audit Plugin shows its own version and the connected devup-mcp version (server-info); hello carries pluginVersion, surfaced in status attachedFiles. Runtime audit adds image-loaded, native-appearance and stale-resources checks. * Verify generated pages with real Chromium rendering and pixel comparisons (#102) * test: require real-browser render verification for generated pages * docs: record CI workflow permission limit --------- Co-authored-by: leader <leader@devfive.invalid> --------- Co-authored-by: leader <leader@devfive.invalid> Co-authored-by: devfive-relay-c70423[bot] <328770141+devfive-relay-c70423[bot]@users.noreply.github.com>
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.
Generated TSX can pass static validation and geometry checks while the page paints incorrectly. This adds an executable real-render gate that runs the existing runtime-audit Playwright runner, then compares each page screenshot against a pinned reference with the production Rust pixel comparator. A successful capture or completed execution alone cannot pass it.
The regression generates a page with production Rust codegen, builds it with React and the actual DevupUI Vite plugin/reset, serves it, and visits it in Chromium. Correct output matches exactly. Missing background and unloaded image each change 93.75% of pixels; the wrong card color changes 6.25%. All four cases retain passing geometry. Tests also cover command exit codes and missing, duplicate and modified references. The harness's normal
npm testcommand runs this end-to-end check along with every existing runtime test.Both devup skills now require browser execution, independently acquired Figma comparisons, report/diff review and retained evidence before claiming completion. Unavailable tooling, missing comparisons and unresolved findings remain incomplete. Updated the embedded skill integrity manifest and generated the patch changepack using the requested bunx command.
Decisions
devfive/3099154426bffbeef8b62999b80bb038/devfive-relay-c70423/devup-mcp-completeness, matching localintegration/devup-mcp-completenessat0c3386a. This keeps the review limited to real-render verification.harness/runtime-audit/verify-render.mjsas a separate command-line gate; the documented MCP capture/ingest/visual-compare/status workflow remains supported.DEVUP_KEEP_RENDER_EVIDENCE=1to retain screenshots, reports and diff PNGs for diagnosis or upload.npm test, but is not automatically gated in repository CI yet. The README includes Linux CI commands; the unapplied workflow patch is retained locally invalidation-ci.patch.log. This limitation is recorded rather than bypassing the GitHub restriction.192.168.0.43:8444for GitHub operations; no edits outside this worktree.Validation
cargo fmt --all --checkcargo clippy --workspace --all-targets -- -D warningscargo test --workspacenpm test --prefix harness/runtime-audit(node --test test/*.test.mjs): 40 passed, no skipsnode --test plugin/tests/*.test.mjs: 8 passedbunx @changepacks/cli --yes --update-type patch --message ...cr-at-eolfor the existing byte-preserved CRLF skill documentsSubmitted via Devfive hub. Employee: ; device:
3099154426bffbeef8b62999b80bb038