fix(ci): exclude projects without previous report from bulk compare by matejchalk · Pull Request #1288 · code-pushup/cli

matejchalk · 2026-09-10T18:17:29Z

In monorepo mode with a bulk run-many command, a project which isn't configured in the base branch (typically a package newly added in the PR) is correctly skipped when collecting previous reports, but it was still included in the subsequent compare command. Since it has no report-before.json, code-pushup compare fails for that project, which fails the whole run-many task and with it the CI job - so no comment is posted, even though every other project compared fine.

The logs of an affected run look like this (Nx monorepo, one new package):

<✓> 14 out of 15 projects loaded previous report from artifact cache
<✓> Switched to base branch main
✖ $ npx nx run @acme/new-pkg:code-pushup -- print-config --output=…
<✓> Executing print-config for project @acme/new-pkg failed, assuming code-pushup not installed in base branch main and skipping comparison
<✓> Skipping 1 projects which aren't configured in base branch main
<✓> Collecting previous reports for 14 projects
<✓> Switched back to PR/MR branch
- $ npx nx run-many --targets=code-pushup --parallel=true --projects=…,@acme/new-pkg,… -- compare
❌ > nx run @acme/new-pkg:code-pushup compare
ENOENT: no such file or directory, open '/…/packages/new-pkg/.code-pushup/report-before.json'
 NX   Running target code-pushup for 15 projects failed

The previous reports are collected with an explicit projects filter (Collecting previous reports for 14 projects), but the compare command called runManyCommand() without any arguments, so the handler fell back to all projects.

Changes:

compareManyProjects passes the names of the projects which actually have a previous report to compareMany, mirroring what collectMany already does for the base branch collect.
Comparison is skipped entirely when no project has a previous report. Previously that case ran compare for every project, none of which had a report-before.json.

As a side effect, the compare command is now always filtered explicitly, so for pnpm/Turborepo/Yarn ≥2 it no longer runs across the whole workspace, but only over the projects which were collected.

npm workspaces and Yarn v1 support no projects filter, so they remain affected by this bug - the same pre-existing limitation applies to the previous report collect.

🤖 Generated with Claude Code

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

pkg-pr-new · 2026-09-10T18:19:40Z

Open in StackBlitz

@code-pushup/ci

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/ci@1288

@code-pushup/cli

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/cli@1288

@code-pushup/core

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/core@1288

@code-pushup/models

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/models@1288

@code-pushup/create-cli

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/create-cli@1288

@code-pushup/axe-plugin

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/axe-plugin@1288

@code-pushup/nx-plugin

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/nx-plugin@1288

@code-pushup/coverage-plugin

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/coverage-plugin@1288

@code-pushup/eslint-plugin

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/eslint-plugin@1288

@code-pushup/js-packages-plugin

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/js-packages-plugin@1288

@code-pushup/jsdocs-plugin

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/jsdocs-plugin@1288

@code-pushup/lighthouse-plugin

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/lighthouse-plugin@1288

@code-pushup/typescript-plugin

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/typescript-plugin@1288

@code-pushup/utils

npm i https://pkg.pr.new/code-pushup/cli/@code-pushup/utils@1288

commit: 5f815aa

github-actions · 2026-09-10T18:37:00Z

Code PushUp

🤨 Code PushUp report has both improvements and regressions – compared current commit 28d014d with previous commit ba41f92.

🕵️ See full comparison in Code PushUp portal 🔍

🏷️ Categories

🏷️ Category	⭐ Previous score	⭐ Current score	🔄 Score change
Performance	🔴 29	🔴 28
Code coverage	🟢 92	🟢 92
Bug prevention	🟡 75	🟡 75	–
Code style	🟢 100	🟢 100	–
Security	🔴 0	🔴 0	–
Updates	🟡 65	🟡 65	–
Type Safety	🟡 67	🟡 67	–
Miscellaneous	🟡 67	🟡 67	–
Documentation	🟡 51	🟡 51	–
Accessibility	🟢 92	🟢 92	–
Best Practices	🟢 97	🟢 97	–
SEO	🟢 92	🟢 92	–
Axe Accessibility	🟡 90	🟡 90	–

👍 1 group improved, 👎 1 group regressed, 👍 2 audits improved, 👎 1 audit regressed, 15 audits changed without impacting score

🗃️ Groups

🔌 Plugin	🗃️ Group	⭐ Previous score	⭐ Current score	🔄 Score change
Lighthouse	Performance	🔴 29	🔴 28
Code coverage	Code coverage metrics	🟢 92	🟢 92

32 other groups are unchanged.

🛡️ Audits

🔌 Plugin	🛡️ Audit	📏 Previous value	📏 Current value
Lighthouse	Total Blocking Time	🟥 1,710 ms	🟥 1,930 ms
Code coverage	Line coverage	🟩 92.3 %	🟩 92.3 %
Code coverage	Branch coverage	🟨 89.2 %	🟨 89.2 %
Lighthouse	Minimizes main-thread work	🟥 8.0 s	🟥 9.3 s
Lighthouse	Largest Contentful Paint	🟥 15.6 s	🟥 14.7 s
Lighthouse	JavaScript execution time	🟥 2.9 s	🟥 3.7 s
Lighthouse	Metrics	🟩 100%	🟩 100%
Lighthouse	Time to Interactive	🟥 16.2 s	🟥 16.6 s
Lighthouse	Avoids enormous network payloads	🟨 Total size was 2,723 KiB	🟨 Total size was 2,723 KiB
Lighthouse	Remove duplicate modules in JavaScript bundles	🟥 Potential savings of 65 KiB	🟥 Potential savings of 68 KiB
Lighthouse	Max Potential First Input Delay	🟥 1,230 ms	🟥 1,050 ms
Lighthouse	Server Backend Latencies	🟩 360 ms	🟩 540 ms
Lighthouse	Reduce unused CSS	🟥 Potential savings of 188 KiB	🟥 Potential savings of 188 KiB
Lighthouse	Initial server response time was short	🟩 Root document took 580 ms	🟩 Root document took 510 ms
Lighthouse	Uses efficient cache policy on static assets	🟨 31 resources found	🟨 31 resources found
Lighthouse	First Contentful Paint	🟥 10.6 s	🟥 10.6 s
Lighthouse	Speed Index	🟥 10.6 s	🟥 10.6 s
Lighthouse	Network Round Trip Times	🟩 50 ms	🟩 50 ms