editorial-standards

# ✍️ Editorial Standards Skill ## 🔴 AI FIRST Quality Principle > **This skill MUST be applied with the AI FIRST principle: never accept first-pass quality. ALL analysis and content MUST go through minimum 2 complete iterations. After first pass, read ALL output back completely and systematically improve every section — strengthen evidence, deepen analysis, add specific citations, broaden perspectives. Spend ALL allocated time on real work. Single-pass output is NEVER acceptable. NO SHORTCUTS.** ## Purpose Provides expertise in maintaining the highest editorial quality standards for political journalism. Covers fact-checking protocols, OSINT/INTOP political intelligence editorial standards, editorial ethics, and quality assurance processes for professional news production. ## Core Principles 1. **Accuracy First** - Every fact must be verifiable and verified 2. **Clarity Always** - Complex ideas explained simply and precisely 3. **Balance Required** - Multiple perspectives fairly represented 4. **Attribution Clear** - Sources properly cited and credited 5. **Style Consistent** - Follow established guidelines rigorously ## This Skill Enforces - **Fact-checking protocols** - Two-source rule, primary source verification - **OSINT/INTOP intelligence style** - Analytical, data-driven, concise, contextual - **AP/Reuters standards** - Accuracy, fairness, transparency - **Copy editing** - Grammar, spelling, punctuation, syntax - **Editorial review** - Multi-stage quality checks before publication - **Error correction** - Transparent, prompt, prominent corrections - **Legal compliance** - Defamation avoidance, privacy protection ## Political Intelligence Editorial Standards ### Writing Principles - **Clarity** - Short sentences, simple words, active voice - **Precision** - Exact meaning, no ambiguity - **Brevity** - No unnecessary words, tight prose - **Elegance** - Sophisticated vocabulary without pretension - **Engagement** - Compelling narratives, reader focus ### Prohibited Practices - ❌ Jargon and buzzwords (unless explained) - ❌ Clichés and overused phrases - ❌ Hyperbole and sensationalism - ❌ Passive voice (except when appropriate) - ❌ Acronyms without definition ### Preferred Style - ✅ Use "said" not "stated" or "claimed" - ✅ Prefer "but" to "however" at sentence start - ✅ Use "will" not "shall" for future tense - ✅ Spell out numbers one to nine, use numerals for 10+ - ✅ Use Oxford comma in lists ## Fact-Checking Protocol ### Pre-Publication Checks **Level 1: Writer Self-Check** - [ ] All facts from authoritative sources - [ ] Two independent sources for major claims - [ ] Direct quotes verbatim and in context - [ ] Statistics from original source documents - [ ] Names, titles, dates verified - [ ] Links/citations functional **Level 2: Copy Editor Review** - [ ] Grammar, spelling, punctuation correct - [ ] Style guide compliance verified - [ ] Logical flow and structure sound - [ ] Technical accuracy confirmed - [ ] Headlines/subheads accurate - [ ] Photo captions correct **Level 3: Fact-Checker Verification** - [ ] All statistics traced to original sources - [ ] Expert claims verified with credentials check - [ ] Historical facts cross-referenced - [ ] Legal claims reviewed by legal expert - [ ] Scientific/technical claims expert-verified - [ ] Potentially controversial claims extra-verified **Level 4: Editorial Review** - [ ] Balanced presentation confirmed - [ ] Ethical standards met - [ ] Legal risk assessed - [ ] Public interest justified - [ ] Headline doesn't overstate - [ ] Final approval for publication ### Source Hierarchy **Primary Sources** (Most Authoritative): 1. Official government documents 2. Legislative records 3. Court filings and judgments 4. Academic peer-reviewed research 5. Statistical agencies (SCB, Eurostat) **Secondary Sources** (Require Verification): 6. News reports from reputable outlets 7. Think tank reports 8. Expert interviews 9. NGO reports 10. Press releases (treated skeptically) **Tertiary Sources** (Generally Avoid): 11. Blog posts 12. Social media posts 13. Wikipedia (useful for leads, not citations) 14. Unverified online sources ## Style Guide Quick Reference ### Editorial Conventions **Dates**: February 6th 2026 (not Feb 6, 2026) **Numbers**: Spell one to nine, use numerals 10+ **Currency**: SEK, euros, dollars (spell out, no symbols in body text) **Percentages**: 10% (use symbol, no space) **Titles**: Lowercase unless part of name (prime minister, but Prime Minister Andersson) **Quotes**: Double quotes "standard", single quotes 'for quotes within quotes' **Commonly Misused Terms**: - Use "which" for non-essential clauses, "that" for essential - "Different from" not "different to" or "different than" - "Between" for two things, "among" for more - "Fewer" for countable, "less" for uncountable - "More than" for numbers, "over" for physical position ### Swedish Political Terminology **Correct Usage**: - Riksdag (not Parliament, except in explanatory context) - Regeringen (Government, not Cabinet) - Myndigheter (Agencies, plural) - Folkpartiet liberalerna (Liberals) → L - Socialdemokraterna (Social Democrats) → S - Moderaterna (Moderates) → M ## Editorial Review Process ### Daily News Cycle **Morning** (08:00-12:00): - News gathering and source development - Initial research and fact-checking - Story assignments and planning **Afternoon** (12:00-17:00): - Writing and initial editing - Fact-checking and verification - Copy editing and style review **Evening** (17:00-20:00): - Editorial review and final approval - Legal review if needed - Multi-language adaptation - Publication scheduling ### Quality Gates **Gate 1: Story Pitch** - Concept approved by editor **Gate 2: Draft Review** - Structure and facts sound **Gate 3: Copy Edit** - Grammar and style correct **Gate 4: Fact Check** - All facts verified **Gate 5: Legal Review** - No defamation/privacy issues **Gate 6: Editorial Approval** - Final green light **Gate 7: Publication** - Multi-language simultaneous release ### Gate 0: Lead-Story & Coverage-Completeness (added 2026-04-18) Before any draft is shared for Gate 2 review, verify: 1. **Lead story** — The article `<title>`, `<meta description>`, Schema.org headline and H1 reference the **#1 finding in `analysis/daily/$DATE/$SUBFOLDER/significance-scoring.md`** under the Democratic-Impact Weighting (DIW) methodology. DIW assigns 30% weight to democratic-infrastructure impact (grundlag, electoral rules, press-freedom law) — this is the highest-priority signal for lead selection. 2. **Lede discipline** — The first two sentences of the lede name the principal human actor (minister, committee chair, party leader) and cite the primary dok_id. 3. **Coverage completeness** — Every document with DIW-weighted score ≥ 7.0 receives a dedicated H3 section in article body. No silent omissions. 4. **Rhetorical tension** — When top-ranked findings carry opposing political valences (e.g., norm entrepreneurship abroad + norm compression at home), an explicit "Rhetorical Cross-Cluster Tension" or equivalent subsection addresses the contradiction. Failure protocol: if any of 1–4 is not satisfied, the draft is returned to the writing agent with the specific missing element identified. **Doctrine**: `analysis/methodologies/ai-driven-analysis-guide.md` §"Rule 5: Democratic-Impact Weighting (DIW) — Lead-Story & Coverage Discipline". **Enforcement**: this is an **editorial / human-review gate**, not a fully automated analysis-gate check. The analysis gate (`.github/prompts/05-analysis-gate.md`) provides partial automation — check 1 (artifact presence of `significance-scoring.md` which drives lead selection) and check 2 (per-document coverage ensuring every manifest `dok_id` has an analysis file a DIW-weighted finding can reference) — but lead-story choice, lede discipline, and article coverage completeness are not machine-enforced here and must be verified by a reviewer before publication. ## Error Correction Protocol ### When Errors Occur **Minor Errors** (spelling, grammar, formatting): - Correct immediately in article - Note in revision history: "Updated: [timestamp] - Minor correction" **Factual Errors** (incorrect data, misattributions): - Correct prominently in article - Add editor's note at top: "Correction: [explanation]" - Update social media posts if shared **Significant Errors** (major facts wrong, unfair representation): - Publish separate correction article - Link from original article prominently - Notify affected parties - Review editorial process to prevent recurrence ## Remember - **Accuracy is non-negotiable** - No compromise for speed - **Style serves clarity** - Rules aid understanding, not hinder - **Political intelligence standard** - Analytical, data-driven, precise - **Multiple checks** - Four-level review process mandatory - **Transparent corrections** - Acknowledge errors promptly, prominently - **Legal awareness** - Defamation, privacy, GDPR compliance - **Source hierarchy** - Primary sources preferred, secondary verified - **Attribution clear** - Readers must know source of every claim - **Balance required** - Multiple perspectives fairly represented - **Copy editing matters** - Grammar errors undermine credibility ## References - [OSINT Framework](https://osintframework.com/) - Open Source Intelligence Tools - [AP Stylebook Online](https://www.apstylebook.com/) - [Reuters Handbook of Journalism](https://www.trust.org/contentAsset/raw-data/652966ab-c90b-4252-b4a5-db8ed1d438ce/file) - [Poynter Institute - Fact-Checking Standards](https://www.poynter.org/) - [International Fact-Checking Network Code of Principles](https://www.poynter.org/ifcn/) - [Swedish Press Ombudsman Editorial Guidelines](https://po.se/) - [European Journalism Centre](https://ejc.net/) --- **Use this skill when**: Writing political news articles, editing submissions for quality and style, fact-checking claims before publication, training journalists on editorial standards, or establishing quality assurance processes for news operations. --- ## 🔗 Integration with agentic workflows & analysis artifacts This skill is consumed by the 11 agentic news workflows in `.github/workflows/news-*.md`. The authoritative contract lives in [`.github/prompts/README.md`](../../prompts/README.md); this skill supplies domain expertise on top of that contract. - **Analysis product** → [`ai-driven-analysis-guide.md`](../../../analysis/methodologies/ai-driven-analysis-guide.md) + every template in [`analysis/templates/`](../../../analysis/templates/). - **Required before any article**: 9 core artifacts (14 for Tier-C) in `analysis/daily/$ARTICLE_DATE/$SUBFOLDER/`; [`05-analysis-gate.md`](../../prompts/05-analysis-gate.md) is the single blocking gate. - **gh-aw v0.69.3** docs: [abridged](https://github.github.com/gh-aw/llms-small.txt) · [complete](https://github.github.com/gh-aw/llms-full.txt) · [blog series](https://github.github.com/gh-aw/_llms-txt/agentic-workflows.txt). ## 📏 Measurable Editorial Quality Bar (CI-enforced) The following metrics are enforced by `scripts/validate-article.ts` (Checks 7–9) as blocking CI checks: | Metric | Threshold | Code | Source | |--------|-----------|------|--------| | **Banned phrases** | 0 hits | `banned-phrase-detected` | [`political-style-guide.json`](../../../analysis/methodologies/political-style-guide.json) | | **Citation density** | ≤ 200 words/anchor (default) | `low-citation-density` | [`reference-quality-thresholds.json`](../../../analysis/methodologies/reference-quality-thresholds.json) → `aiFirst.citationDensity.perArticle` | | **economicProvenance vintage** | ≤ 6 months or annotated | `stale-economic-provenance` | [ECONOMIC_DATA_CONTRACT.md](../../aw/ECONOMIC_DATA_CONTRACT.md) | | **Pass-2 attestation** | mtime > birth + 180s | Check 6 in gate | `05-analysis-gate.md` | ### Pass-2 improvement metrics (tracked in `methodology-reflection.md`) - Net word-count change: ≥ 5% between Pass-1 and Pass-2 artifacts - Citation count delta: must increase - Banned-phrase count: must reach 0 by Pass-2 end