Overview

7 days30 days

Latest CI Pipeline Executions

Fuzzy

Failed
36636
0bc9ee9e fix(core): apply package.json updates unblocked by later package groups
by Leosvel Pé...
Canceled
36636
88097f66 fix(core): apply package.json updates unblocked by later package groups
by Leosvel Pé...
Succeeded
master
c75960a2 fix(devkit): correct version floor validation for ranges and prereleases (#36724) ## Current Behavior Generators calling devkit's `assertSupportedPackageVersion` reject a workspace whose package is declared as a version range, even when the installed version is supported. The declared range is coerced to its first version token and compared to the floor: ``` [@nx/js:typescript-sync]: Unsupported version of `typescript` detected. Installed: >=4.8.4 <6.1.0 Supported: >= 5.8.0 ``` This workspace has TypeScript 6.0.2 installed (a pnpm override), which satisfies the floor. ## Expected Behavior The installed version decides when it satisfies the declared range, so the workspace above passes validation. A declared range that cannot reach the floor still fails as unsupported. A range that straddles the floor with no resolvable installed version fails asking to install dependencies first. Prereleases count as their release version. ## Related Issue(s) NXC-4820 ## Implementation Notes - The installed version is read from the tree's node_modules first, then through workspace-root module resolution for layouts without node_modules (Yarn PnP). Resolution is restricted to the workspace root so a copy under `.nx/installation` cannot shadow the workspace's install. - A straddling range without a resolvable install cannot be judged: the lockfile may pin either side, and `.nx/installation` setups run generators before workspace packages are installed. It fails with an install-dependencies error rather than guessing in either direction. - An installed prerelease matches the declared range in raw form (same-tuple prerelease comparator) or as its release version; the floor comparison always uses the release version.  --- <a href="https://app.trypolygraph.com/orgs/6a061dcb561c062131116eca/sessions/nxc-4820-7fbd7f9d">View Polygraph session ↗</a>  --------- Co-authored-by: nx-cloud[bot] <71083854+nx-cloud[bot]@users.noreply.github.com>
by Leosvel Pé...
Failed
36724
Flaky task→
4d341573 cleanup(devkit): use readJson to read the installed package version from the tree
by Leosvel Pé...
Succeeded
36642
cca55a7c cleanup(core): tighten the daemon environment spec The negative getDaemonSpawnEnv assertion passed even without the undefined guard (an undefined-valued property still satisfies toBeUndefined), so assert the key is absent instead. The convergence test failed for developers with JAVA_TOOL_OPTIONS exported because the captured payload carried the shell value; delete it before capturing. Move the MISE_SHELL assertion to the shell-session test to match the grouping in the exclusion list.
by Leosvel Pé...
Succeeded
master
a7ba17c4 chore(repo): derive exact PR deltas after base movement (#36720) ## Current Behavior The rebase guard added in #36707 infers base movement from changed-file counts and compares normalized per-file patches. Base-branch context churn can surface as author work, path iteration can split valid filenames, and genuine deltas below 40 lines are skipped. ## Expected Behavior Re-reviews determine base movement from the prior and current merge bases. When the base moves, the previously reviewed patch is replayed onto the current base and compared with HEAD. Only an exactly empty author delta skips the agent pass. Replay conflicts fall back to the full PR diff, while Git and infrastructure failures remain fatal.  --- <a href="https://app.trypolygraph.com/orgs/6a061dcb561c062131116eca/sessions/review-pr-rebase-guard-follow-up-9a4f2742">View Polygraph session ↗</a> 
by Leosvel Pé...
Succeeded
master
a7ba17c4 chore(repo): derive exact PR deltas after base movement (#36720) ## Current Behavior The rebase guard added in #36707 infers base movement from changed-file counts and compares normalized per-file patches. Base-branch context churn can surface as author work, path iteration can split valid filenames, and genuine deltas below 40 lines are skipped. ## Expected Behavior Re-reviews determine base movement from the prior and current merge bases. When the base moves, the previously reviewed patch is replayed onto the current base and compared with HEAD. Only an exactly empty author delta skips the agent pass. Replay conflicts fall back to the full PR diff, while Git and infrastructure failures remain fatal.  --- <a href="https://app.trypolygraph.com/orgs/6a061dcb561c062131116eca/sessions/review-pr-rebase-guard-follow-up-9a4f2742">View Polygraph session ↗</a> 
by Leosvel Pé...
Failed
36724
Flaky task→
4d341573 cleanup(devkit): use readJson to read the installed package version from the tree
by Leosvel Pé...
Succeeded
master
a7ba17c4 chore(repo): derive exact PR deltas after base movement (#36720) ## Current Behavior The rebase guard added in #36707 infers base movement from changed-file counts and compares normalized per-file patches. Base-branch context churn can surface as author work, path iteration can split valid filenames, and genuine deltas below 40 lines are skipped. ## Expected Behavior Re-reviews determine base movement from the prior and current merge bases. When the base moves, the previously reviewed patch is replayed onto the current base and compared with HEAD. Only an exactly empty author delta skips the agent pass. Replay conflicts fall back to the full PR diff, while Git and infrastructure failures remain fatal.  --- <a href="https://app.trypolygraph.com/orgs/6a061dcb561c062131116eca/sessions/review-pr-rebase-guard-follow-up-9a4f2742">View Polygraph session ↗</a> 
by Leosvel Pé...
Succeeded
master
a7ba17c4 chore(repo): derive exact PR deltas after base movement (#36720) ## Current Behavior The rebase guard added in #36707 infers base movement from changed-file counts and compares normalized per-file patches. Base-branch context churn can surface as author work, path iteration can split valid filenames, and genuine deltas below 40 lines are skipped. ## Expected Behavior Re-reviews determine base movement from the prior and current merge bases. When the base moves, the previously reviewed patch is replayed onto the current base and compared with HEAD. Only an exactly empty author delta skips the agent pass. Replay conflicts fall back to the full PR diff, while Git and infrastructure failures remain fatal.  --- <a href="https://app.trypolygraph.com/orgs/6a061dcb561c062131116eca/sessions/review-pr-rebase-guard-follow-up-9a4f2742">View Polygraph session ↗</a> 
by Leosvel Pé...
Succeeded
36724
b431a3fc cleanup(devkit): correct getInstalledPackageVersion JSDoc for generator-time fallback
by Leosvel Pé...
Succeeded
master
a7ba17c4 chore(repo): derive exact PR deltas after base movement (#36720) ## Current Behavior The rebase guard added in #36707 infers base movement from changed-file counts and compares normalized per-file patches. Base-branch context churn can surface as author work, path iteration can split valid filenames, and genuine deltas below 40 lines are skipped. ## Expected Behavior Re-reviews determine base movement from the prior and current merge bases. When the base moves, the previously reviewed patch is replayed onto the current base and compared with HEAD. Only an exactly empty author delta skips the agent pass. Replay conflicts fall back to the full PR diff, while Git and infrastructure failures remain fatal.  --- <a href="https://app.trypolygraph.com/orgs/6a061dcb561c062131116eca/sessions/review-pr-rebase-guard-follow-up-9a4f2742">View Polygraph session ↗</a> 
by Leosvel Pé...
Failed
36724
Flaky task→
b431a3fc cleanup(devkit): correct getInstalledPackageVersion JSDoc for generator-time fallback
by Leosvel Pé...
Succeeded
master
a7ba17c4 chore(repo): derive exact PR deltas after base movement (#36720) ## Current Behavior The rebase guard added in #36707 infers base movement from changed-file counts and compares normalized per-file patches. Base-branch context churn can surface as author work, path iteration can split valid filenames, and genuine deltas below 40 lines are skipped. ## Expected Behavior Re-reviews determine base movement from the prior and current merge bases. When the base moves, the previously reviewed patch is replayed onto the current base and compared with HEAD. Only an exactly empty author delta skips the agent pass. Replay conflicts fall back to the full PR diff, while Git and infrastructure failures remain fatal.  --- <a href="https://app.trypolygraph.com/orgs/6a061dcb561c062131116eca/sessions/review-pr-rebase-guard-follow-up-9a4f2742">View Polygraph session ↗</a> 
by Leosvel Pé...
Canceled
36547
5febdc8f chore(repo): merge master into feat/convert-to-inferred-single-inference
by Leosvel Pé...
Succeeded
36724
dc4f8f31 fix(devkit): correct version floor validation for ranges and prereleases Generators calling assertSupportedPackageVersion rejected workspaces declaring a version range that a supported installed version satisfies. `@nx/js:typescript-sync` reported an unsupported typescript version for a declared range of ">=4.8.4 <6.1.0" with TypeScript 6.0.2 installed, because the declared range was coerced to its first version token (4.8.4) and compared to the 5.8.0 floor. The assertion now resolves the installed version first and lets it decide when it satisfies the declared range. It reads the tree's node_modules and falls back to workspace-root module resolution for installs without a node_modules layout, e.g. Yarn PnP. Prereleases count as their release version throughout. A declared range that cannot reach the floor still fails as unsupported. A range that straddles the floor without a resolvable install fails asking to install dependencies first, since the lockfile may pin either side; this is reachable with .nx/installation setups where plugins run before workspace packages are installed.
by Leosvel Pé...
Succeeded
36402
97017898 docs(testing): cover ESM configs and lowercase proxy variables The working-directory workaround said to derive addresses from __dirname, but the generated playwright.config.mts is ESM, where that identifier does not exist; the page now names import.meta.dirname for ESM configs and __dirname for CommonJS ones. The probe variable list only had the uppercase spellings, while the lowercase ones also count and win when both are set.
by Leosvel Pé...
Succeeded
master
35b83b49 cleanup(angular): bump e2e es2015 bundle size threshold for Angular 22.1.3 (#36723) ## Current Behavior The `e2e-angular` bundle size assertion fails deterministically on master and PR branches: ``` The current es2015 bundle size is 226.238 KB expect(es2015BundleSize).toBeLessThanOrEqual(226000); // fails ``` Angular 22.1.3 was published on 2026-08-19 and is resolved via the `~22.1.0` pin in `packages/angular/src/utils/versions.ts`. It ships larger fesm2022 code than 22.1.2: platform-browser +494 B, core +119 B, router +14 B. That pushes the test app's es2015 bundle to 226,238 bytes, past the 226,000-byte threshold. ## Expected Behavior The bundle size threshold is 227,000 bytes, which accommodates the Angular 22.1.3 bundle and unblocks CI on master. ## Bundle growth attribution - Rebuilding an equivalent app (webpack bundler, `RouterModule.forRoot`) with each version pinned reproduces the growth: `main.js` goes from 226,180 to 226,698 bytes (+518 B). - Bumping one Angular package at a time splits that delta exactly: router +349 B, platform-browser +159 B, core +10 B. - The entire delta is Angular runtime code shipped in 22.1.3; nothing on the Nx side contributes to the growth.  --- <a href="https://app.trypolygraph.com/orgs/6a061dcb561c062131116eca/sessions/bump-angular-bundle-threshold-bb7a6285">View Polygraph session ↗</a> 
by Leosvel Pé...
Succeeded
36723
ebd7aa5e cleanup(angular): bump e2e es2015 bundle size threshold for Angular 22.1.3 Angular 22.1.3 ships larger fesm2022 bundles than 22.1.2 (platform-browser +494 B, core +119 B, router +14 B), pushing the e2e app's es2015 bundle to 226,238 bytes and deterministically failing the 226,000-byte assertion on master and PR branches. Raise the threshold to 227,000 bytes.
by Leosvel Pé...
Failed
36547
Flaky task→
5febdc8f chore(repo): merge master into feat/convert-to-inferred-single-inference
by Leosvel Pé...