@cryptotaxi247 / SquadScope-1 / commits / 2a00eb1

fix(seo): close pipeline gaps in meta descriptions and cross-linking (#522)

* fix(seo): close pipeline gaps in meta descriptions and cross-linking - Add summary field to yearly frontmatter derived from narrative - Add 155-char max constraint for weekly summary in analysis prompt - Add max 70-char guidance for weekly title in analysis prompt - Add cross-links: monthly→yearly and monthly→weekly navigation - Add cross-links: yearly→monthly navigation in rendered pages - Monthly pages already have summary populated (verified) Closes #521 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * fix: address all 5 Copilot review comments on PR #522 - Fix _extract_summary regex to handle .!? terminators with whitespace lookahead and strip leading whitespace - Rename unused 'month' param to '_month' in _build_monthly_crosslinks - Move summary constraint from sub-bullet to inline rule 13 text in analyze-weekly.md prompt - Move nav-link injection into build_yearly_narrative_pages so both render_yearly_page and generate_rollups pipelines include monthly navigation links - Add _extract_summary tests covering sentence terminators, truncation, word boundaries, whitespace, and frontmatter integration Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Juan Manuel Servera committed Jun 17, 2026 at 00:53 UTC 2a00eb196b22a89d7709a2bc570cdcf6477006af

4 files changed +125 -6

prompts/analyze-weekly.md

+3 -3

@@ -127,7 +127,7 @@ Be critical, selective, and opinionated.

127 - `quality_score`

128 - `summary`

129 Optional: `predictions`

130 -6. `title` must be a punchy 5-12 word journalistic headline that captures the week's dominant themes. Never use generic week/year labels such as `Week NN, YYYY Analysis` or `Week NN, YYYY`.

130 +6. `title` must be a punchy 5-12 word journalistic headline (max 70 characters) that captures the week's dominant themes. Never use generic week/year labels such as `Week NN, YYYY Analysis` or `Week NN, YYYY`.

131 - Good: `Agent Skills, Exploit Churn, and the Language Nobody Asked For`

132 - Good: `The Week Local Models Went Mainstream`

133 - Good: `MCP Eats the Middleware Layer While VCs Look Elsewhere`

@@ -137,7 +137,7 @@ Be critical, selective, and opinionated.

137 10. `repos_featured` should equal the total number of repos considered in the weekly editorial pass.

138 11. `stars_tracked` should equal the total stars across those repos.

139 12. `top_repo` should be the repo that best anchors the editorial narrative, not automatically the most-starred repo.

140 -13. `quality_score` must be an honest 0-100 self-assessment; publishable work is `>= 60`.

140 +13. `quality_score` must be an honest 0-100 self-assessment; publishable work is `>= 60`. The `summary` field must be ≤155 characters, a complete sentence crafted as the meta description for search engines and social sharing. Do not let it exceed 155 characters.

141 14. If you include `predictions`, each entry must be `{repo, claim_type, direction, confidence}` with `claim_type` in `signal|noise|gap`, `direction` in `up|flat|down`, and `confidence` from `0` to `1`.

142 15. Include all required sections in this exact order:

143

@@ -199,7 +199,7 @@ repos_featured: 0

199 stars_tracked: 0

200 top_repo: "owner/repo"

201 quality_score: 0

202 -summary: "One-sentence editorial thesis."

202 +summary: "One-sentence editorial thesis (max 155 chars, used as meta description)."

203 predictions:

204 - repo: owner/repo

205 claim_type: signal

scripts/generate_rollups.py

+11 -1

@@ -355,6 +355,15 @@ def monthly_entries(weekly: WeeklySummary, tags_counter: Counter[str]) -> dict[s

355 }

356

357

358 +def _build_monthly_crosslinks(year: int, _month: int, items: list[WeeklySummary]) -> str:

359 + """Build navigation cross-links for a monthly page."""

360 + links = []

361 + links.append(f"[{year} Year in Review](/yearly/{year}/)")

362 + for item in items:

363 + links.append(f"[{item.week_title}]({item.week_link})")

364 + return f"*Part of {links[0]}* · Weekly: {' · '.join(links[1:])}\n"

365 +

366 +

367 def build_monthly_pages(summaries: list[WeeklySummary], content_root: Path, analyzed_dir: Path) -> list[RollupPage]:

368 grouped: dict[tuple[int, int], list[WeeklySummary]] = defaultdict(list)

369 for summary in summaries:

@@ -368,7 +377,8 @@ def build_monthly_pages(summaries: list[WeeklySummary], content_root: Path, anal

377

378 synthesis = ensure_month_synthesis(items, analyzed_dir)

379

371 - synthesis_text = synthesis.narrative

380 + crosslinks = _build_monthly_crosslinks(year, month, items)

381 + synthesis_text = crosslinks + "\n" + synthesis.narrative

382 if synthesis.trend_arc:

383 synthesis_text += f"\n\n### Trend Arc\n\n{synthesis.trend_arc}"

384 page_entries["Month Synthesis"].append(

scripts/generate_yearly_narrative.py

+35 -2

@@ -680,6 +680,25 @@ def generate_yearly_title(year: int, arcs: dict[str, list[str]]) -> str:

680 return title[:70]

681

682

683 +def _extract_summary(narrative: str, max_length: int = 155) -> str:

684 + """Extract a ≤155-character summary from the narrative's first sentence."""

685 + narrative = narrative.strip()

686 + # Take first sentence (up to first . ! or ? followed by whitespace or end)

687 + first_sentence_match = re.match(r"(.+?[.!?])(?:\s|$)", narrative)

688 + if first_sentence_match:

689 + sentence = first_sentence_match.group(1).strip()

690 + if len(sentence) <= max_length:

691 + return sentence

692 + # Truncate at last word boundary within limit

693 + truncated = sentence[:max_length - 1].rsplit(" ", 1)[0]

694 + return truncated.rstrip(".,;:") + "…"

695 + # Fallback: truncate narrative at word boundary

696 + if len(narrative) <= max_length:

697 + return narrative.strip()

698 + truncated = narrative[:max_length - 1].rsplit(" ", 1)[0]

699 + return truncated.rstrip(".,;:") + "…"

700 +

701 +

702 def build_yearly_narrative_pages(content_root: Path, years: Iterable[int] | None = None) -> list[YearlyNarrativePage]:

703 grouped: dict[int, list[MonthSnapshot]] = {}

704 for snapshot in load_month_snapshots(content_root, years):

@@ -691,6 +710,19 @@ def build_yearly_narrative_pages(content_root: Path, years: Iterable[int] | None

710 narrative = synthesize_year(ordered)

711 arcs = {family.key: detect_family_arc(ordered, family) for family in TREND_FAMILIES}

712 title = generate_yearly_title(year, arcs)

713 + summary = _extract_summary(narrative)

714 + month_slugs = [month.month_slug for month in ordered]

715 + nav_links = []

716 + for slug in month_slugs:

717 + parts = slug.split("-")

718 + if len(parts) == 2:

719 + year_str, month_str = parts

720 + month_num = int(month_str)

721 + month_name = MONTH_NAMES.get(month_num, month_str)

722 + nav_links.append(f"[{month_name}](/monthly/{year_str}/{month_str}/)")

723 + nav_prefix = ""

724 + if nav_links:

725 + nav_prefix = f"**Monthly reports:** {' · '.join(nav_links)}\n\n"

726 pages.append(

727 YearlyNarrativePage(

728 year=year,

@@ -700,10 +732,11 @@ def build_yearly_narrative_pages(content_root: Path, years: Iterable[int] | None

732 "date": ordered[-1].date,

733 "year": year,

734 "categories": ["yearly"],

703 - "months_covered": [month.month_slug for month in ordered],

735 + "months_covered": month_slugs,

736 "format": "narrative",

737 + "summary": summary,

738 },

706 - narrative=narrative,

739 + narrative=nav_prefix + narrative,

740 )

741 )

742 return pages

tests/test_generate_rollups.py

+76

@@ -490,6 +490,82 @@ The strongest thread was a shift from raw capability talk toward packaging, trus

490 self.assertEqual(velocity["old-tag"], "dying")

491 self.assertIn(velocity["new-thing"], ("accelerating", "new"))

492

493 + def test_extract_summary_in_yearly_frontmatter(self) -> None:

494 + """Yearly pages must include a summary field in frontmatter."""

495 + with temporary_workspace() as tmpdir:

496 + base = Path(tmpdir)

497 + analyzed_dir = base / "data" / "analyzed"

498 + content_root = base / "content"

499 + analyzed_dir.mkdir(parents=True)

500 +

501 + (analyzed_dir / "2026-W21-summary.md").write_text(

502 + make_summary(

503 + week="2026-W21",

504 + date="2026-05-18T12:07:20+00:00",

505 + top_repo="octo/signal-kit",

506 + summary="Practical agent tooling led the week.",

507 + signal="Teams preferred operational automation over generic hype.",

508 + noise="Exploit-heavy projects still added editorial noise.",

509 + gaps="Reliable momentum data remained missing.",

510 + conclusion="The strongest projects made automation safer to adopt.",

511 + ),

512 + encoding="utf-8",

513 + )

514 +

515 + generate_rollups.generate_rollups(analyzed_dir, content_root)

516 + yearly_path = content_root / "yearly" / "2026.md"

517 + yearly = yearly_path.read_text(encoding="utf-8")

518 + self.assertIn("summary:", yearly)

519 + # Extract summary value and verify length constraint

520 + for line in yearly.splitlines():

521 + if line.strip().startswith("summary:"):

522 + summary_value = line.split(":", 1)[1].strip().strip('"')

523 + self.assertLessEqual(len(summary_value), 155)

524 + break

525 + else:

526 + self.fail("summary field not found in yearly frontmatter")

527 +

528 +

529 +class ExtractSummaryTests(unittest.TestCase):

530 + def test_short_sentence_returned_as_is(self) -> None:

531 + result = generate_yearly_narrative._extract_summary("Short sentence.")

532 + self.assertEqual(result, "Short sentence.")

533 +

534 + def test_exclamation_mark_terminates_sentence(self) -> None:

535 + result = generate_yearly_narrative._extract_summary("Wow! More text follows here.")

536 + self.assertEqual(result, "Wow!")

537 +

538 + def test_question_mark_terminates_sentence(self) -> None:

539 + result = generate_yearly_narrative._extract_summary("Why not? The rest is irrelevant.")

540 + self.assertEqual(result, "Why not?")

541 +

542 + def test_truncation_respects_max_length(self) -> None:

543 + long = "A" * 200 + "."

544 + result = generate_yearly_narrative._extract_summary(long, max_length=50)

545 + self.assertLessEqual(len(result), 50)

546 + self.assertTrue(result.endswith("…"))

547 +

548 + def test_truncation_at_word_boundary(self) -> None:

549 + sentence = "This is a moderately long sentence that should be truncated at a word boundary when it exceeds the maximum allowed length for meta descriptions."

550 + result = generate_yearly_narrative._extract_summary(sentence, max_length=60)

551 + self.assertLessEqual(len(result), 60)

552 + self.assertTrue(result.endswith("…"))

553 + self.assertFalse(result[-2].isspace())

554 +

555 + def test_leading_whitespace_stripped(self) -> None:

556 + result = generate_yearly_narrative._extract_summary(" Leading spaces. More text.")

557 + self.assertEqual(result, "Leading spaces.")

558 +

559 + def test_fallback_when_no_sentence_terminator(self) -> None:

560 + result = generate_yearly_narrative._extract_summary("No punctuation at all")

561 + self.assertEqual(result, "No punctuation at all")

562 +

563 + def test_fallback_truncation_for_long_text_without_terminator(self) -> None:

564 + long = "word " * 50

565 + result = generate_yearly_narrative._extract_summary(long, max_length=30)

566 + self.assertLessEqual(len(result), 30)

567 + self.assertTrue(result.endswith("…"))

568 +

569

570 if __name__ == "__main__":

571 unittest.main()