Add end-to-end integration tests for news article generation pipeline by Copilot · Pull Request #707 · Hack23/riksdagsmonitor

Copilot · 2026-03-01T01:53:55Z

Original prompt

This section details on the original issue you should resolve

<issue_title>Add end-to-end integration tests for news article generation pipeline</issue_title>
<issue_description>## 📋 Issue Type
Architecture / Quality

🎯 Objective

Add end-to-end tests for the news article generation pipeline, validating the full flow from template rendering through article quality enhancement to final HTML output.

📊 Current State

53 test files with 2169 tests pass

Tests cover individual modules (template, quality enhancer, MCP client, etc.)

No integration test validates the full article generation pipeline end-to-end

News generation is only validated after deployment via validate-news-generation.sh

Critical path from data → template → HTML → validation is tested in isolation, not as a pipeline

🚀 Desired State

Integration tests for the full news generation pipeline

Tests cover: data fetching mock → content generation → template rendering → HTML validation → quality checks

Tests verify all 14 language variants produce valid output

Tests verify Schema.org structured data is correct

Tests verify hreflang tags are consistent

Tests run as part of CI (not just post-deployment)

🔧 Implementation Approach

Create tests/pipeline/news-generation-e2e.test.ts

Mock MCP client responses with realistic data

Run full pipeline: generateArticleHTML() → validate with HTMLHint → check quality metrics

Verify output for each article type:

week-ahead, month-ahead, weekly-review, monthly-review

committee-reports, propositions, motions

evening-analysis

Test edge cases:

Empty data (no parliamentary activity)

Missing translations

RTL language output (Arabic, Hebrew)

Very long article content

Verify Schema.org JSON-LD validity

🤖 Recommended Agent

test-specialist — Integration test development

✅ Acceptance Criteria

Pipeline integration tests for all 8 article types

Tests cover all 14 language variants

HTMLHint validation built into tests

Schema.org JSON-LD validated

hreflang consistency verified

Tests run in CI via npm test

Edge cases covered (empty data, RTL, long content)

📚 References

Pipeline tests: tests/pipeline/ (existing directory)

Article template: scripts/article-template/template.ts

Quality enhancer: scripts/article-quality-enhancer.ts

News types: scripts/news-types/ (8 article type modules)

Test config: vitest.config.js

🏷️ Labels

type:testing, priority:medium, component:news-generation</issue_description>

<agent_instructions>Add end-to-end tests for the news generation pipeline in tests/pipeline/news-generation-e2e.test.ts. Mock MCP client, run full pipeline through generateArticleHTML(), validate with HTMLHint patterns, check Schema.org JSON-LD. Cover all 8 article types and edge cases.</agent_instructions>

Comments on the Issue (you are @copilot in this section)

Fixes Add end-to-end integration tests for news article generation pipeline #688

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

Co-authored-by: pethers <1726836+pethers@users.noreply.github.com>

github-actions · 2026-03-01T03:02:54Z

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

Copilot

Pull request overview

Adds an end-to-end Vitest integration test suite that exercises the news article generation pipeline as a single flow (mock MCP data → generator → HTML template output → pipeline HTML validation), improving confidence beyond the existing unit-level coverage.

Changes:

Adds tests/pipeline/news-generation-e2e.test.ts with 41 tests covering all 8 news-type generators and validating generated HTML via pipeline/validation.
Verifies all 14 language variants for correct <html lang="…">, RTL dir="rtl" behavior, and structural validity.
Adds assertions for Schema.org JSON-LD parseability/shape and hreflang alternate-link consistency.

Copilot · 2026-03-01T03:36:29Z

tests/pipeline/news-generation-e2e.test.ts

+  it('contains hreflang tags for all 14 language codes', () => {
+    const html = articleTemplate.generateArticleHTML(makeArticleData('en'));
+    ALL_14_LANGS.forEach(lang => {
+      expect(html, `hreflang="${lang}" missing`).toContain(`hreflang="${lang}"`);


The hreflang assertions treat Norwegian as hreflang="no", but elsewhere in the repo hreflang is explicitly mapped to BCP-47 nb for Norwegian (no filename suffix → nb hreflang). These tests should follow that convention (expect hreflang="nb" while keeping the href ending in -no.html) so the E2E suite doesn’t codify an SEO inconsistency.

Suggested change

expect(html, `hreflang="${lang}" missing`).toContain(`hreflang="${lang}"`);

const expectedHreflang = lang === 'no' ? 'nb' : lang;

expect(html, `hreflang="${expectedHreflang}" missing`).toContain(`hreflang="${expectedHreflang}"`);

Fixed in commit 9e976b9. The template (scripts/article-template/template.ts line 112) now maps no → nb in the hreflang attribute (URL path stays -no.html), consistent with generate-sitemap.ts, generate-rss.ts, and generate-sitemap-html.ts. The test now checks hreflang="nb" for Norwegian.

Copilot · 2026-03-01T03:36:29Z

tests/pipeline/news-generation-e2e.test.ts

+    // Generator should still succeed (degraded article or error, but not throw)
+    expect(typeof result.success).toBe('boolean');
+  });
+
+  it('committee-reports handles empty reports gracefully', async () => {
+    mockClientInstance.fetchCommitteeReports.mockResolvedValue([]);
+
+    const result = await committeeReportsMod.generateCommitteeReports({ languages: ['en'] });
+    expect(typeof result.success).toBe('boolean');
+  });
+
+  it('propositions handles empty data gracefully', async () => {
+    mockClientInstance.fetchPropositions.mockResolvedValue([]);
+
+    const result = await propositionsMod.generatePropositions({ languages: ['en'] });
+    expect(typeof result.success).toBe('boolean');
+  });
+
+  it('motions handles empty data gracefully', async () => {
+    mockClientInstance.fetchMotions.mockResolvedValue([]);
+
+    const result = await motionsMod.generateMotions({ languages: ['en'] });
+    expect(typeof result.success).toBe('boolean');


The “empty MCP data” edge-case tests only assert that result.success is a boolean, which will still pass even if a generator regresses into always failing (or returning success without generating any HTML). To meaningfully validate graceful degradation, assert the expected outcome per generator (e.g., success: true with files: 0 for “skip” generators, or success: true + articles[0].html passing validateArticleHTML for generators that still emit a degraded article).

Suggested change

// Generator should still succeed (degraded article or error, but not throw)

expect(typeof result.success).toBe('boolean');

});

it('committee-reports handles empty reports gracefully', async () => {

mockClientInstance.fetchCommitteeReports.mockResolvedValue([]);

const result = await committeeReportsMod.generateCommitteeReports({ languages: ['en'] });

expect(typeof result.success).toBe('boolean');

});

it('propositions handles empty data gracefully', async () => {

mockClientInstance.fetchPropositions.mockResolvedValue([]);

const result = await propositionsMod.generatePropositions({ languages: ['en'] });

expect(typeof result.success).toBe('boolean');

});

it('motions handles empty data gracefully', async () => {

mockClientInstance.fetchMotions.mockResolvedValue([]);

const result = await motionsMod.generateMotions({ languages: ['en'] });

expect(typeof result.success).toBe('boolean');

// Generator should still succeed and emit a degraded but valid article

expect(result.success).toBe(true);

const articles = result.articles as GeneratedArticle[];

expect(Array.isArray(articles)).toBe(true);

expect(articles.length).toBeGreaterThan(0);

const validation = pipelineValidation.validateArticleHTML(articles[0]!.html);

expect(validation.passed, validation.errors.join(', ')).toBe(true);

});

it('committee-reports handles empty reports gracefully', async () => {

mockClientInstance.fetchCommitteeReports.mockResolvedValue([]);

const result = await committeeReportsMod.generateCommitteeReports({ languages: ['en'] });

expect(result.success).toBe(true);

const articles = result.articles as GeneratedArticle[];

expect(Array.isArray(articles)).toBe(true);

expect(articles.length).toBeGreaterThan(0);

const validation = pipelineValidation.validateArticleHTML(articles[0]!.html);

expect(validation.passed, validation.errors.join(', ')).toBe(true);

});

it('propositions handles empty data gracefully', async () => {

mockClientInstance.fetchPropositions.mockResolvedValue([]);

const result = await propositionsMod.generatePropositions({ languages: ['en'] });

expect(result.success).toBe(true);

const articles = result.articles as GeneratedArticle[];

expect(Array.isArray(articles)).toBe(true);

expect(articles.length).toBeGreaterThan(0);

const validation = pipelineValidation.validateArticleHTML(articles[0]!.html);

expect(validation.passed, validation.errors.join(', ')).toBe(true);

});

it('motions handles empty data gracefully', async () => {

mockClientInstance.fetchMotions.mockResolvedValue([]);

const result = await motionsMod.generateMotions({ languages: ['en'] });

expect(result.success).toBe(true);

const articles = result.articles as GeneratedArticle[];

expect(Array.isArray(articles)).toBe(true);

expect(articles.length).toBeGreaterThan(0);

const validation = pipelineValidation.validateArticleHTML(articles[0]!.html);

expect(validation.passed, validation.errors.join(', ')).toBe(true);

Strengthened in commit 9e976b9. Each generator now has specific assertions:

week-ahead (empty events): success: true + articles.length > 0 + HTML passes validateArticleHTML (with requireSections: false since no events → no <h2> sections is expected)

committee-reports, propositions, motions (empty data): success: true + files: 0 (they skip article generation when no source data is available)

pethers · 2026-03-01T03:40:44Z

@copilot apply changes based on the comments in this thread

github-actions · 2026-03-01T03:42:08Z

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

…case assertions Co-authored-by: pethers <1726836+pethers@users.noreply.github.com>

github-actions · 2026-03-01T10:57:35Z

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

Copilot

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

Copilot · 2026-03-01T11:25:52Z

scripts/article-template/template.ts

  <!-- Hreflang for language alternatives -->
-${ALL_LANG_CODES.map(l => `  <link rel="alternate" hreflang="${l}" href="https://riksdagsmonitor.com/news/${baseSlug}-${l}.html">`).join('\n')}
+${ALL_LANG_CODES.map(l => `  <link rel="alternate" hreflang="${l === 'no' ? 'nb' : l}" href="https://riksdagsmonitor.com/news/${baseSlug}-${l}.html">`).join('\n')}
  <link rel="alternate" hreflang="x-default" href="https://riksdagsmonitor.com/news/${baseSlug}-en.html">


The <link rel="alternate"> hreflang mapping now emits nb for Norwegian, but the page still renders the language switcher links with hreflang="no" via generateArticleLanguageSwitcher() (helpers.ts). This creates inconsistent hreflang values for the same Norwegian variant within one document. Consider centralizing a hreflangCode() mapping (no→nb) in the article-template layer and using it for both the <link rel="alternate"> tags and the language-switcher <a hreflang> attributes (or removing hreflang from the switcher entirely if it’s not needed).

Fixed in commit 10d1947. Added a hreflangCode() helper to helpers.ts that centralizes the no → nb BCP-47 mapping. Both generateArticleLanguageSwitcher() (the in-page switcher) and the <link rel="alternate"> tags in template.ts now use this shared helper, ensuring consistent hreflang="nb" for Norwegian across the entire document.

Copilot · 2026-03-01T11:25:53Z

tests/pipeline/news-generation-e2e.test.ts

+  it('contains hreflang tags for all 14 language codes', () => {
+    const html = articleTemplate.generateArticleHTML(makeArticleData('en'));
+    ALL_14_LANGS.forEach(lang => {
+      const expectedHreflang = lang === 'no' ? 'nb' : lang;
+      expect(html, `hreflang="${expectedHreflang}" missing`).toContain(`hreflang="${expectedHreflang}"`);
+    });
+  });


The hreflang assertions only validate the <link rel="alternate"> tags. Since the template also emits hreflang attributes in the in-page language switcher, this suite won’t catch the current mismatch where Norwegian renders as hreflang="no" in the switcher but hreflang="nb" in the head alternates. Add an assertion that the language-switcher link for the -no.html variant uses hreflang="nb" (or whatever the intended mapping is) to prevent regressions.

Added in commit 10d1947. A new test 'language switcher Norwegian link uses hreflang="nb" (BCP-47), not "no"' asserts that no anchor with -no.html href uses hreflang="no", and that hreflang="nb" is present in the switcher markup.

pethers · 2026-03-01T11:42:03Z

@copilot apply changes based on the comments in this thread

github-actions · 2026-03-01T11:43:24Z

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

github-actions · 2026-03-01T13:44:10Z

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

Copilot

Pull request overview

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

Comments suppressed due to low confidence (1)

tests/pipeline/news-generation-e2e.test.ts:345

ltrLangs is a subset of LTR languages, but the test name says it covers “LTR languages”. As written, a regression that incorrectly adds dir="rtl" to e.g. da, no, fi, es, or nl would not be caught. Consider deriving this list from ALL_14_LANGS (exclude only ar/he) so the assertion matches the intent.

  it('does not set dir="rtl" for LTR languages', () => {
    const ltrLangs: Language[] = ['en', 'sv', 'de', 'fr', 'ja', 'ko', 'zh'];
    ltrLangs.forEach(lang => {
      const html = articleTemplate.generateArticleHTML(makeArticleData(lang));
      expect(html, `${lang} should not have rtl`).not.toContain('dir="rtl"');
    });
  });

Copilot · 2026-03-01T13:45:42Z

tests/pipeline/news-generation-e2e.test.ts

+    const results = pipelineValidation.validateArticleBatch(articles);
+
+    results.forEach(r => {
+      expect(r.passed, `Language ${r.filename} failed: ${r.errors.join(', ')}`).toBe(true);
+    });
+  });


In this test, validateArticleBatch(articles) is only iterated via results.forEach(...). If validateArticleBatch regresses to return an empty array, the loop won’t run and the test will still pass. Add an assertion that results.length === articles.length (or toHaveLength(articles.length)) before iterating so the batch validator itself is actually validated.

Fixed in commit 2250320. Added expect(results).toHaveLength(articles.length) before the forEach so a regressed validateArticleBatch returning an empty array will now fail fast.

Copilot · 2026-03-01T13:45:42Z

tests/pipeline/news-generation-e2e.test.ts

+    const successful = results.filter(r => r.success && r.articles?.length);
+    expect(successful.length).toBeGreaterThan(0);
+
+    const batch = successful.map(r => {


This test is titled “across all 8 article types”, but it filters to successful and only asserts successful.length > 0. That means the test can still pass even if several generators start failing (or stop producing articles), which defeats the purpose of covering all 8 types. Consider asserting each results[i].success is true and that you have 8 generated articles in the batch before calling validateArticleBatch.

Suggested change

const successful = results.filter(r => r.success && r.articles?.length);

expect(successful.length).toBeGreaterThan(0);

const batch = successful.map(r => {

// Ensure all 8 generators succeeded and produced at least one article each

expect(results).toHaveLength(8);

results.forEach((r, index) => {

expect(r.success, `Generator index ${index} failed`).toBe(true);

expect(

r.articles && r.articles.length > 0,

`Generator index ${index} produced no articles`,

).toBe(true);

});

const batch = results.map(r => {

Fixed in commit 2250320. Replaced the filter + successful.length > 0 pattern with explicit assertions: expect(results).toHaveLength(8), per-generator expect(r.success).toBe(true), and expect(r.articles && r.articles.length > 0).toBe(true) before building the batch.

pethers · 2026-03-01T13:54:12Z

@copilot apply changes based on the comments in this thread and TypeScript & JavaScript Testing / Unit Tests (Vitest) (pull_request)
TypeScript & JavaScript Testing / Unit Tests (Vitest) (pull_request)Failing after 1m

github-actions · 2026-03-01T13:55:24Z

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

… all 8 generators succeed Co-authored-by: pethers <1726836+pethers@users.noreply.github.com>

Copilot · 2026-03-01T14:03:31Z

@copilot apply changes based on the comments in this thread

Applied in commit 2250320. Both batch-validation weaknesses are now fixed: the "all 14 language variants" test guards against an empty-array regression with a length check, and the "all 8 article types" test explicitly asserts all 8 generators succeed and produce articles before building the batch.

github-actions · 2026-03-01T14:15:04Z

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

Copilot

Pull request overview

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

Comments suppressed due to low confidence (1)

tests/pipeline/news-generation-e2e.test.ts:305

This test file hardcodes the full 14-language list in ALL_14_LANGS even though the template layer already exposes a single source of truth (ALL_LANG_CODES). To avoid the E2E suite drifting if languages are added/removed/reordered, consider importing and using ALL_LANG_CODES directly instead of duplicating the list here.

// All 14 supported language codes (mirrors ALL_LANG_CODES from constants)
// ---------------------------------------------------------------------------

const ALL_14_LANGS: readonly Language[] = [
  'en', 'sv', 'da', 'no', 'fi', 'de', 'fr', 'es', 'nl', 'ar', 'he', 'ja', 'ko', 'zh',
];

Copilot · 2026-03-01T14:18:12Z

tests/pipeline/news-generation-e2e.test.ts

+    );
+    expect(html).toContain('<!DOCTYPE html>');


The test name says it validates handling of HTML entities/special characters, but it doesn’t assert that potentially unsafe characters (like </>) are escaped in the rendered HTML. With the current inputs (subtitle contains <key>), the test will still pass even if the template outputs raw HTML, so it won’t catch XSS/markup-injection regressions. Consider asserting that <key> is rendered as <key> (or, if raw HTML is intended, rename the test and avoid implying escaping).

Suggested change

);

expect(html).toContain('<!DOCTYPE html>');

);

// Basic HTML structure check

expect(html).toContain('<!DOCTYPE html>');

// Ensure potentially unsafe characters are HTML-escaped in the rendered output.

// The raw `<key>` sequence should not appear, but its escaped form should.

expect(html).not.toContain('Analysis of <key> policy areas & legislative priorities');

expect(html).toContain('Analysis of <key> policy areas & legislative priorities');

// JSON-LD blocks should still be valid

Fixed in commit 921a709. Renamed the test to "generates valid HTML when title and subtitle contain special characters" (removes the misleading "HTML entities" implication). Removed <key> from the subtitle (the template renders the lede raw by design — escaping is applied to meta tags and JSON-LD, as covered by article-template.test.ts:511-522). Added expect(html).toContain('Budget & Defence') to assert that & IS escaped in the meta/JSON-LD context.

pethers · 2026-03-01T14:23:05Z

@copilot apply changes based on the comments in this thread and TypeScript & JavaScript Testing / Unit Tests (Vitest) (pull_request)
TypeScript & JavaScript Testing / Unit Tests (Vitest) (pull_request)Failing after 1m

…pecial-chars e2e test Co-authored-by: pethers <1726836+pethers@users.noreply.github.com>

Copilot · 2026-03-01T14:32:41Z

@copilot apply changes based on the comments in this thread and [TypeScript & JavaScript Testing / Unit Tests (Vitest...

Fixed in commit 921a709. Two issues addressed:

CI failure (tests/fix-article-navigation.test.ts): updated buildCompleteArticle fixture to hreflang="nb" for Norwegian and added the no → nb mapping to the "generates all 14 language links" assertion — matches the hreflangCode() BCP-47 mapping now used by generateArticleLanguageSwitcher.
Review comment (special-chars test): renamed test to not imply body-level escaping; removed <key> from subtitle; added & assertion for meta/JSON-LD context.

github-actions · 2026-03-01T14:39:41Z

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

Copilot

Pull request overview

Copilot reviewed 5 out of 5 changed files in this pull request and generated no new comments.

Initial plan

95c5d9d

Copilot AI assigned Copilot and pethers Mar 1, 2026

Copilot AI mentioned this pull request Mar 1, 2026

Comprehensive codebase analysis: 18 issues created and assigned to Copilot #671

Draft

Copilot started work on behalf of pethers March 1, 2026 02:49 View session

Add end-to-end integration tests for news generation pipeline

f19d384

Co-authored-by: pethers <1726836+pethers@users.noreply.github.com>

Copilot AI changed the title ~~[WIP] Add end-to-end integration tests for news article generation pipeline~~ Add end-to-end integration tests for news article generation pipeline Mar 1, 2026

Copilot finished work on behalf of pethers March 1, 2026 03:00

Merge branch 'main' into copilot/add-integration-tests-news-pipeline

e5f1d81

github-actions bot added testing Test coverage size-l Large change (250-1000 lines) labels Mar 1, 2026

pethers requested a review from Copilot March 1, 2026 03:30

Copilot started reviewing on behalf of pethers March 1, 2026 03:31 View session

Copilot AI reviewed Mar 1, 2026

View reviewed changes

Merge branch 'main' into copilot/add-integration-tests-news-pipeline

f7e8aab

Copilot started work on behalf of pethers March 1, 2026 03:40 View session

Fix Norwegian hreflang (no→nb BCP-47) and strengthen empty-data edge …

9e976b9

…case assertions Co-authored-by: pethers <1726836+pethers@users.noreply.github.com>

Copilot finished work on behalf of pethers March 1, 2026 03:48

Merge branch 'main' into copilot/add-integration-tests-news-pipeline

4db5174

github-actions bot added the refactor Code refactoring label Mar 1, 2026

pethers requested a review from Copilot March 1, 2026 11:16

Copilot started reviewing on behalf of pethers March 1, 2026 11:17 View session

Copilot AI reviewed Mar 1, 2026

View reviewed changes

Merge branch 'main' into copilot/add-integration-tests-news-pipeline

155fb62

pethers requested a review from Copilot March 1, 2026 13:42

Copilot started reviewing on behalf of pethers March 1, 2026 13:43 View session

Copilot AI reviewed Mar 1, 2026

View reviewed changes

Merge branch 'main' into copilot/add-integration-tests-news-pipeline

4f63640

Copilot started work on behalf of pethers March 1, 2026 13:57 View session

fix: strengthen batch validation assertions — length guard and assert…

2250320

… all 8 generators succeed Co-authored-by: pethers <1726836+pethers@users.noreply.github.com>

Copilot finished work on behalf of pethers March 1, 2026 14:04

Merge branch 'main' into copilot/add-integration-tests-news-pipeline

e194c5d

pethers requested a review from Copilot March 1, 2026 14:13

Copilot started reviewing on behalf of pethers March 1, 2026 14:14 View session

Copilot AI reviewed Mar 1, 2026

View reviewed changes

Copilot started work on behalf of pethers March 1, 2026 14:23 View session

fix: update fix-article-navigation tests for hreflang nb and rename s…

921a709

…pecial-chars e2e test Co-authored-by: pethers <1726836+pethers@users.noreply.github.com>

Copilot finished work on behalf of pethers March 1, 2026 14:33

Merge branch 'main' into copilot/add-integration-tests-news-pipeline

7da7e21

pethers requested a review from Copilot March 1, 2026 14:38

Copilot started reviewing on behalf of pethers March 1, 2026 14:39 View session

Copilot AI reviewed Mar 1, 2026

View reviewed changes

pethers approved these changes Mar 1, 2026

View reviewed changes

pethers marked this pull request as ready for review March 1, 2026 14:44

pethers merged commit b0c473a into main Mar 1, 2026
19 checks passed

pethers deleted the copilot/add-integration-tests-news-pipeline branch March 1, 2026 14:44

	expect(html, `hreflang="${lang}" missing`).toContain(`hreflang="${lang}"`);
	const expectedHreflang = lang === 'no' ? 'nb' : lang;
	expect(html, `hreflang="${expectedHreflang}" missing`).toContain(`hreflang="${expectedHreflang}"`);

-    const successful = results.filter(r => r.success && r.articles?.length);
-    expect(successful.length).toBeGreaterThan(0);
-    const batch = successful.map(r => {
+    // Ensure all 8 generators succeeded and produced at least one article each
+    expect(results).toHaveLength(8);
+    results.forEach((r, index) => {
+      expect(r.success, `Generator index ${index} failed`).toBe(true);
+      expect(
+        r.articles && r.articles.length > 0,
+        `Generator index ${index} produced no articles`,
+      ).toBe(true);
+    });
+    const batch = results.map(r => {

-    );
-    expect(html).toContain('<!DOCTYPE html>');
+    );
+    // Basic HTML structure check
+    expect(html).toContain('<!DOCTYPE html>');
+    // Ensure potentially unsafe characters are HTML-escaped in the rendered output.
+    // The raw `<key>` sequence should not appear, but its escaped form should.
+    expect(html).not.toContain('Analysis of <key> policy areas & legislative priorities');
+    expect(html).toContain('Analysis of &lt;key&gt; policy areas &amp; legislative priorities');
+    // JSON-LD blocks should still be valid

Conversation

Copilot AI commented Mar 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🎯 Objective

📊 Current State

🚀 Desired State

🔧 Implementation Approach

🤖 Recommended Agent

✅ Acceptance Criteria

📚 References

🏷️ Labels

Comments on the Issue (you are @copilot in this section)

Uh oh!

github-actions bot commented Mar 1, 2026

🔍 Lighthouse Performance Audit

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

pethers commented Mar 1, 2026

Uh oh!

github-actions bot commented Mar 1, 2026

🔍 Lighthouse Performance Audit

Uh oh!

github-actions bot commented Mar 1, 2026

🔍 Lighthouse Performance Audit

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

pethers commented Mar 1, 2026

Uh oh!

github-actions bot commented Mar 1, 2026

🔍 Lighthouse Performance Audit

Uh oh!

github-actions bot commented Mar 1, 2026

🔍 Lighthouse Performance Audit

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI commented Mar 1, 2026 •

edited

Loading