📙 Strategic Extensions Methodology

📊 Family C — Depth Layer for High-Significance Events
🎯 Scenario Analysis · Comparative International · Devil's Advocate (ACH) · Intelligence Assessment · Methodology Reflection

📋 Document Owner: CEO | 📄 Version: 1.4 | 📅 Last Updated: 2026-05-15 (UTC) 🔄 Review Cycle: Quarterly | ⏰ Next Review: 2026-07-21 🏢 Owner: Hack23 AB (Org.nr 5595347807) | 🏷️ Classification: Public

---

🎯 AI-FIRST Methodology Card

🚦 Read this card before writing a single paragraph. It names the artifact this methodology owns, the gate check it satisfies, the evidence-density target it must hit, and the Pass-1 / Pass-2 discipline required by .github/copilot-instructions.md §5 (AI-FIRST Quality Principle).

Field	Value
Purpose	Family C — step-by-step production of the 5 strategic-extension artifacts (scenarios, comparative international, ACH devil's advocate, intelligence assessment, methodology reflection).
Inputs	Family A synthesis + SWOT + risk + threat; OSINT tradecraft standards; reference-quality-thresholds for the article type
Outputs	scenario-analysis.md, comparative-international.md, devils-advocate.md, intelligence-assessment.md, methodology-reflection.md
Owning artifact(s)	All 5 Family C artifacts
Owning gate check	Check 7 (Family C structure: ≥ 3 scenarios, ≥ 3 KJs with confidence, ≥ 3 ACH hypotheses, ICD 203 audit marker)
Citation density target	≥ 1 evidence anchor per scenario/hypothesis/KJ; ≥ 2 cycle-aged sources per year/cycle horizon claim
Banned phrases	Enforced via political-style-guide.md §Machine-readable banned-phrase list
Threshold source	reference-quality-thresholds.json → thresholds[articleType][artifact] (fallback defaults.coreArtifactFloor)

✅ Pass-1 checklist (creation — minimal viable artifact)

• ≥ 3 distinct scenarios with WEP probability and total ≤ 100%
• ≥ 3 ACH competing hypotheses with explicit consistent/inconsistent/N-A scoring per evidence row
• ≥ 3 Key Judgments each with VERY HIGH/HIGH/MEDIUM/LOW/VERY LOW confidence and ≥ 1 PIR reference
• methodology-reflection.md contains an ICD 203 audit marker or ≥ 3 named methodology improvements
• Produce every required sub-section listed in the owning template
• Add ≥ 1 evidence anchor (dok_id, vote id, named MP, or primary-source URL) per analytical claim
• Apply the correct WEP confidence band for the run's horizon (72h / week / month / quarter / year / cycle)
• Include ≥ 1 themed Mermaid diagram with style … or themeVariables config (where structurally meaningful)
• Cross-link the relevant template under analysis/templates/ and the gate check it satisfies

🔁 Pass-2 checklist (read-back & improve — AI-FIRST mandatory)

• Add ≥ 1 wildcard / black-swan to scenario-analysis where horizon ≥ year
• Verify comparator set covers ≥ 2 peer countries with named indicators (IMF / SCB / WB)
• Read back devils-advocate: each hypothesis must survive at least one elimination test
• Re-read the file end-to-end; flag every claim that lacks an evidence anchor and add one
• Replace every banned phrase listed in political-style-guide.md §Machine-readable banned-phrase list with an evidence-anchored alternative
• Tighten WEP language: never above likely without ≥ 3 cycle-aged sources for year/cycle horizons
• Strengthen Mermaid (color-coded style … directives, themeVariables, ≥ 5 nodes where the structure admits it)
• Add ≥ 1 second-order effect, cui-bono note, or counterfactual where the artifact admits one
• Verify citation density meets the per-file target below and the gate's evidence-density rules

🟢 Exemplar (good — pattern-match this)

(scenario row) "Scenario B — Tidö coalition holds (likely, ~70 %): M+KD+L+SD pass H902FiU1 budget; precedent: 2024 budget vote (cf. H801FiU1, 175–174). Wildcard: SD walkout if migration deal slips (~10 %). Sources: dok_id H902FiU1, statskontoret.se freshness 2026-04-22."

🔴 Anti-exemplar (failure mode — never ship this)

(failure mode) "Several scenarios are possible. The coalition might hold or might not. Experts believe the situation is fluid." — < 3 scenarios, no probabilities, banned phrases, no dok_id.

🔗 Cross-links

• Template(s): analysis/templates/scenario-analysis.md, analysis/templates/comparative-international.md, analysis/templates/devils-advocate.md, analysis/templates/intelligence-assessment.md, analysis/templates/methodology-reflection.md
• Gate check: .github/prompts/05-analysis-gate.md
• AI-FIRST canon: .github/copilot-instructions.md §5 · ai-driven-analysis-guide.md
• Style canon: political-style-guide.md · osint-tradecraft-standards.md
• Catalog row: artifact-catalog.md

---

🔄 Tradecraft Anchors

Element	Value	Reference
F3EAD Stage	ANALYZE	This methodology covers deep analytical processing — competing hypotheses, alternative futures, cross-country comparison
PIRs Served	Per-file: intelligence-assessment.md declares served PIRs; devils-advocate challenges PIR-relevant hypotheses; comparative-international benchmarks PIR-4 (Defence), PIR-5 (Fiscal)	See political-style-guide.md §PIR/EEI Catalog
Admiralty Floor	devils-advocate.md requires ≥[B2] evidence per hypothesis; intelligence-assessment.md Key Judgments require ≥[A1] or ≥2×[B2]	See political-style-guide.md §Admiralty Code
WEP Requirement	scenario-analysis.md probabilities in WEP language; intelligence-assessment.md Key Judgments with WEP + ODNI confidence	See political-style-guide.md §WEP + ODNI
ICD 203 Gate	Standard 2 (uncertainties), 3 (judgments vs assumptions), 4 (alternative analysis), 7 (explain changes)	See political-style-guide.md §ICD 203
SAT(s)	ACH, Red Team, Devil's Advocacy (devils-advocate.md); What If?, Morphological (scenario-analysis.md); Outside-In Thinking (comparative-international.md); Key Assumptions Check, Quality of Information Check (methodology-reflection.md)	See political-style-guide.md §SATs

---

🎯 Purpose

Family C delivers analytic depth when the daily event set warrants it. Where Family A narrates what happened, Family C answers:

• What could happen next? (scenario-analysis)
• How does this compare internationally? (comparative-international)
• What are we getting wrong? (devils-advocate — ACH)
• What does this mean at the intelligence level? (intelligence-assessment)
• Was our process sound? (methodology-reflection)

Core — every run produces all 5

Family C files are always produced on every workflow run. They are not trigger-driven. Depth per file adapts to the day's DIW distribution, but the output set is stable: every folder ships scenario-analysis.md, comparative-international.md, devils-advocate.md, intelligence-assessment.md, and methodology-reflection.md.

🧭 Family C is exactly 5 files — not 6. executive-brief.md is Family A (synthesis layer), not Family C. If a downstream document or README enumerates 6 Family C files including executive-brief.md, that document is stale — the canonical inventory is artifact-catalog.md §"Family C — Strategic Extensions (5 artifacts)". methodology-reflection.md belongs to Family C even though its production order (Step 5, after the other four) and its self-audit role make it feel meta — the gate at 05-analysis-gate.md check 7 enforces this exact 5-file structure.

File	Behaviour on a light news day	Behaviour on a P0-dense day
scenario-analysis.md	Three scenarios converge on a narrow-band consensus; file documents why branching is low	Three divergent scenarios with probability-weighted indicators
comparative-international.md	Compares baseline Swedish position to ≥5 Nordic + EU peers	Peer-country reform response with deep causal analysis
devils-advocate.md	ACH on the day's #1 ranked document with ≥3 competing hypotheses	Full ACH on every P0 + red-team hypothesis + Bayesian base-rate check
intelligence-assessment.md	3 Key Judgments tied to PIRs for the next 72 h	5–7 Key Judgments with confidence + warning indicators + HUMINT/OSINT cross-checks
⭐ methodology-reflection.md	Vital run-audit gate. Evidence sufficiency, confidence distribution, source diversity, party-neutrality arithmetic, ≥3 concrete methodology improvements for the next cycle	Same structure; skipping the file breaks the self-correction loop and fails the quality gate

⭐ methodology-reflection.md is vital. It is the only file that audits the run itself. Treat a missing or stub methodology-reflection.md as a broken workflow and revise before commit. The Pass-2 rewrite in the AI-Driven Guide Step 7 reads this file before revising every other file.

flowchart LR
    classDef famA fill:#E8F5E9,stroke:#4CAF50,color:#1B5E20
    classDef scenario fill:#E3F2FD,stroke:#1565C0,color:#0D47A1
    classDef ach fill:#FFEBEE,stroke:#D32F2F,color:#B71C1C
    classDef intl fill:#FFF8E1,stroke:#FFC107,color:#3E2723
    classDef intel fill:#F3E5F5,stroke:#7B1FA2,color:#311B92
    classDef reflect fill:#FFF3E0,stroke:#FF9800,color:#BF360C,stroke-width:3px

    A[Family A<br/>synthesis + impact]:::famA
    C1[scenario-analysis.md<br/>🔮 plausible futures]:::scenario
    C2[comparative-international.md<br/>🌍 EU/Nordic parallels]:::intl
    C3[devils-advocate.md<br/>⚔️ ACH + Red Team]:::ach
    C4[intelligence-assessment.md<br/>🎯 strategic bottom line]:::intel
    C5[⭐ methodology-reflection.md<br/>🔬 VITAL run-audit gate]:::reflect

    A --> C1
    A --> C2
    A --> C3
    C1 --> C4
    C3 --> C4
    C2 --> C4
    C4 --> C5

---

🔮 Part 1 — Scenario Analysis (scenario-analysis.md)

Purpose

Explore ≥3 plausible forward paths for the top P0/P1 items over a 30–180 day horizon, each evidenced and each with pre-declared indicators so analysts can track which path materialises.

Input

• significance-scoring.md (which items to scenario)
• synthesis-summary.md (current narrative baseline)
• stakeholder-perspectives.md (actor positions that drive branching)
• Historical baseline (Family D historical-parallels.md when available)

Output — required structure

1. Baseline (expected) — the path implied by current stated positions + prior revealed preferences
2. Upside — a plausible path more favourable to democratic accountability / institutional strength
3. Downside — a plausible path more hostile to the status quo
4. Wildcard — low-probability / high-impact branch (≤10 % probability)
5. For each scenario:
   • Narrative (≤120 words)
   • Triggering events (≥3 concrete, each with a source or dated leading indicator)
   • Blocking events (what would make this scenario fail)
   • Probability (with 5-level confidence label)
   • Consequence summary (actor-by-actor, institution-by-institution)
   • Early-warning indicators (observable within 14 days)
6. Scenario-probability Mermaid — color-coded probability bar
7. Branching decision Mermaid — color-coded flowchart of key forks
8. Cross-scenario comparison table — impact × probability × reversibility

Required Mermaid — branching paths

flowchart TD
    classDef base fill:#9E9E9E,stroke:#424242,color:#FFFFFF
    classDef upside fill:#4CAF50,stroke:#1B5E20,color:#FFFFFF
    classDef downside fill:#FF9800,stroke:#E65100,color:#FFFFFF
    classDef wild fill:#D32F2F,stroke:#B71C1C,color:#FFFFFF
    classDef fork fill:#FFC107,stroke:#F57F17,color:#3E2723

    T0[Today — P0 document filed]:::base
    F1{Fork 1<br/>committee vote ≤14d}:::fork
    F2{Fork 2<br/>opposition cohesion}:::fork

    S_base[Baseline<br/>orderly passage · 55 %]:::base
    S_up[Upside<br/>compromise amendment · 20 %]:::upside
    S_down[Downside<br/>coalition fracture · 20 %]:::downside
    S_wild[Wildcard<br/>snap extra utskott · 5 %]:::wild

    T0 --> F1
    F1 -->|pass| F2
    F1 -->|block| S_down
    F2 -->|unified| S_base
    F2 -->|split| S_up
    F1 -->|escalation| S_wild

Quality gate

• Probabilities sum to 100 %
• Each scenario has ≥3 named triggering events with sources
• Each scenario has ≥2 early-warning indicators with ISO dates
• Wildcard probability ≤10 %
• Blocking events named and sourced
• Cross-scenario table present

---

🌍 Part 2 — Comparative International (comparative-international.md)

Filename variant: international-comparative.md — identical structure.

Purpose

Place the Swedish political event in international context so readers understand precedent, best practice, and likely interactions with EU / Nordic / OECD peers.

Input

• synthesis-summary.md (current events)
• IMF WEO / FM / IFS / BOP / GFS_COFOG / DOTS / PCPS / MFS_IR / ER (economic context); SCB for Swedish-specific ground truth; World Bank for governance (WGI source=75), environment, social/education participation, defence historicals, crime/justice
• EU legislative database references (where EU law intersects)
• Named peer countries (default set: DK, NO, FI, DE, NL, EE for Nordic+Baltic-West benchmark)

Output — required structure

1. Issue framing — ≤80 words: what is the Swedish question being compared?
2. Peer-country evidence table — for each peer country:
   • Country · Approach (brief) · Outcome (quantified) · Source · Applicability to Sweden
3. Best-practice extraction — three elements worth importing, with sources
4. Incompatibility notes — three elements that do not travel well, with reasons
5. EU-law intersection — directives, regulations, and open infringement procedures that apply
6. Comparative Mermaid — color-coded country grid on the chosen axis (e.g. policy permissiveness)
7. Benchmark trend chart — quantitative time-series Mermaid using IMF macro/fiscal data as primary economic source (WB WGI / environment / social residue may be charted only where the axis is non-economic — governance, environment, social participation)

Required Mermaid — peer grid

graph LR
    classDef high fill:#4CAF50,stroke:#1B5E20,color:#FFFFFF
    classDef med fill:#FFC107,stroke:#F57F17,color:#3E2723
    classDef low fill:#FF9800,stroke:#E65100,color:#FFFFFF
    classDef crit fill:#D32F2F,stroke:#B71C1C,color:#FFFFFF
    classDef sv fill:#1565C0,stroke:#0D47A1,color:#FFFFFF

    SE[Sweden<br/>current SE-specific score]:::sv
    DK[Denmark<br/>high-strictness regime]:::high
    NO[Norway<br/>medium-strictness regime]:::med
    FI[Finland<br/>medium-strictness regime]:::med
    DE[Germany<br/>low-strictness regime]:::low
    NL[Netherlands<br/>low-strictness regime]:::low
    EE[Estonia<br/>critical-outlier datapoint]:::crit

    SE -.benchmark.- DK
    SE -.benchmark.- NO
    SE -.benchmark.- FI
    SE -.benchmark.- DE
    SE -.benchmark.- NL
    SE -.benchmark.- EE

Quality gate

• ≥5 peer countries included (default Nordic+Baltic-West set is acceptable)
• Every peer row has a quantified outcome and a source URL
• Applicability column distinguishes constitutional, institutional, and operational transferability
• EU-law intersection lists specific directive/regulation numbers
• Benchmark chart included when IMF economic data exists, or World Bank non-economic data exists for governance, environment, or social axes

---

⚔️ Part 3 — Devil's Advocate (devils-advocate.md)

Purpose

Apply the Analysis of Competing Hypotheses (ACH) technique to stress-test the dominant interpretation of the day's events. Surfaces alternative explanations, red-teams the narrative, and quantifies residual uncertainty.

Input

• synthesis-summary.md (candidate narrative)
• stakeholder-perspectives.md (declared positions)
• Family B cross-reference-map.md (coordinated-activity patterns)
• Open-source intelligence (OSINT) on actor history

Output — required structure

1. Dominant hypothesis (H1) — the synthesis's main interpretation, stated in ≤40 words
2. Alternative hypotheses (H2, H3, … Hn, n ≥ 3) — each a genuinely different interpretation
3. Evidence matrix — rows: observable evidence items; columns: each hypothesis; cells: Supports (+), Contradicts (−), Ambiguous (~)
4. Diagnostic value analysis — each piece of evidence scored by how much it discriminates between hypotheses
5. Residual uncertainty statement — what would have to be true to flip the ranking
6. Red-team section — a maximally adversarial interpretation from the POV of a hostile actor
7. Confidence summary — per-hypothesis confidence label
8. ACH Mermaid — color-coded outcome network

Required Mermaid — ACH outcome

graph TB
    classDef win fill:#4CAF50,stroke:#1B5E20,color:#FFFFFF
    classDef lose fill:#D32F2F,stroke:#B71C1C,color:#FFFFFF
    classDef neutral fill:#9E9E9E,stroke:#424242,color:#FFFFFF
    classDef red fill:#C2185B,stroke:#880E4F,color:#FFFFFF

    H1[H1 — dominant<br/>policy continuity<br/>🟩 HIGH confidence]:::win
    H2[H2 — alternative<br/>pre-election signalling<br/>🟧 MEDIUM]:::neutral
    H3[H3 — alternative<br/>coalition renegotiation<br/>🟧 MEDIUM]:::neutral
    H4[H4 — red-team<br/>deliberate misdirection<br/>🟥 LOW]:::red
    H5[H5 — rejected<br/>uncoordinated noise<br/>⬛ VERY LOW]:::lose

    Evidence[Observable evidence set] --> H1
    Evidence --> H2
    Evidence --> H3
    Evidence --> H4
    Evidence --> H5
``$

### \text{ACH} \text{scoring} \text{rules} — \text{positive}-\text{voice}
- \text{Produce} \text{an} \text{evidence} \text{matrix} \text{with} \text{at} \text{least} 5 \text{rows} \text{of} \text{diagnostic} \text{evidence}
- \text{Rank} \text{hypotheses} \text{by} \text{inconsistency} \text{score} (\text{fewer} − \text{cells} \text{wins})
- \text{Call} \text{out} \text{the} \text{evidence} \text{item} \text{whose} \text{removal} \text{would} \text{flip} \text{the} \text{top}-2 \text{ranking} — \text{that}'\text{s} \text{the} "\text{fragility} \text{point}"
- \text{Include} \text{a} \text{hypothesis} \text{that} \text{a} \text{hostile} \text{actor} \text{would} \text{push} (\text{Red} \text{Team} \text{H\_n})
- \text{Pre}-\text{declare} \text{a} \text{disconfirming} \text{observation} \text{that} \text{would} \text{reverse} \text{the} \text{assessment}

### \text{Quality} \text{gate}
- [ ] \text{n} ≥ 3 \text{alternative} \text{hypotheses}, \text{each} \text{substantively} \text{different}
- [ ] \text{Evidence} \text{matrix} ≥5 \text{rows}  \times  ≥4 \text{hypothesis} \text{columns}
- [ ] \text{Fragility} \text{point} \text{identified}
- [ ] \text{Red}-\text{team} \text{hypothesis} \text{present} \text{with} \text{explicit} "\text{this} \text{is} \text{adversarial} \text{framing}" \text{label}
- [ ] \text{Residual} \text{uncertainty} \text{statement} \text{quantified}

---

## 🎯 \text{Part} 4 — \text{Intelligence} \text{Assessment} ($intelligence-assessment.md`)

### Purpose
Deliver the **strategic bottom line** at intelligence-community quality: single assessment paragraph per question, confidence label, alternative view, and collection gap.

### Input
- synthesis-summary.md · stakeholder-impact.md · scenario-analysis.md · devils-advocate.md
- Behavioral-analysis insights (actor history, cognitive biases)
- Prior intelligence-assessment.md files from the last 30 days (for continuity)

### Output — required structure

1. **Key Judgments (KJs)** — 3–7 numbered paragraphs, each with:
   - Assertion (≤60 words)
   - Confidence label (5-level scale)
   - Key evidence (≥2 citations)
   - Alternative view (one sentence)
2. **Strategic implications** — 3 bullets on what this means for decision-makers over 30–180 days
3. **Collection gaps** — 3 explicit questions the evidence base does not answer
4. **Priority intelligence requirements (PIR)** — 3 things the next cycle should prioritise
5. **Tradecraft note** — which analytic techniques were applied (ACH, SWOT, PESTLE, Network, etc.)
6. **Intelligence Mermaid** — color-coded confidence bar per KJ

### Required Mermaid — KJ confidence

```mermaid
graph LR
    classDef vh fill:#1565C0,stroke:#0D47A1,color:#FFFFFF
    classDef h fill:#4CAF50,stroke:#1B5E20,color:#FFFFFF
    classDef m fill:#FFC107,stroke:#F57F17,color:#3E2723
    classDef l fill:#FF9800,stroke:#E65100,color:#FFFFFF
    classDef vl fill:#D32F2F,stroke:#B71C1C,color:#FFFFFF

    KJ1[KJ1 — coalition holds<br/>🟦 VERY HIGH]:::vh
    KJ2[KJ2 — budget amendment passes<br/>🟩 HIGH]:::h
    KJ3[KJ3 — opposition unity weakens<br/>🟧 MEDIUM]:::m
    KJ4[KJ4 — EU infringement risk<br/>🟥 LOW]:::l
    KJ5[KJ5 — snap election trigger<br/>⬛ VERY LOW]:::vl

Quality gate

• 3–7 Key Judgments
• Confidence labels distributed realistically (not all HIGH)
• Every KJ has ≥2 citations
• 3 collection gaps named
• 3 PIRs formatted as answerable questions
• Tradecraft note lists techniques explicitly

---

🔬 Part 5 — Methodology Reflection (methodology-reflection.md)

Purpose

Transparent analytic audit of the workflow run — what worked, what didn't, what the next cycle must fix. This file is the platform's self-correction mechanism.

Input

• All other outputs from the workflow
• Comparison with prior day's equivalent outputs
• Tool-call logs from the workflow runner
• Confidence-label distribution across outputs

Output — required structure

1. Evidence sufficiency audit — for each Family A output, was evidence sufficient? (yes/no with reason)
2. Confidence distribution — histogram Mermaid of confidence labels across the workflow
3. Source diversity — percentage of claims from Riksdag API vs Regeringen vs SCB vs international
4. Bias audit — were all 8 parties given fair analytical depth given their evidence footprint? Quantify.
5. Tradecraft audit — which techniques were applied; which were skipped and why
6. Forward methodology adjustments — 3 concrete improvements for next cycle, each actionable
7. Prior-period comparison — what has the methodology done better / worse vs last workflow?

Required Mermaid — confidence distribution

pie showData
    title Confidence label distribution across this workflow
    "🟦 VERY HIGH" : 12
    "🟩 HIGH" : 34
    "🟧 MEDIUM" : 38
    "🟥 LOW" : 12
    "⬛ VERY LOW" : 4

Positive-voice bias audit rule

Produce a table of parties × claims-about-them × depth-score (words + citations) so a reader can see neutrality arithmetically. Acceptable outcome: no party's depth deviates from the evidence-weighted average by more than 25 %.

Quality gate

• Evidence sufficiency assessed per Family A output
• Confidence distribution Mermaid present
• Source diversity percentages sum to 100 %
• Bias audit quantified with specific numbers
• ≥3 concrete methodology adjustments named for next cycle

---

🌐 Long-horizon extensions (horizonDays ≥ 90)

DRY policy. This section summarizes operational floors; canonical values live in the bounded-context authority:

• Authoritative source for horizon stratification, scenario-tree depth, counterfactual minima, IMF projection-year stamps, PESTLE-mandatory thresholds, cross-horizon citation rules and forward-indicator bands → .github/prompts/ext/long-horizon-forecasting.md.
• Aggregation multipliers (1.7 / 2.0 / 2.5) and sibling-folder ingestion → .github/prompts/ext/tier-c-aggregation.md.
• Gate enforcement → .github/prompts/05-analysis-gate.md, check LH-3 = counterfactual paragraph minima.

Long-horizon scenario-tree composition

When the run resolves to quarter-ahead, year-ahead or election-cycle, scenario-analysis.md MUST branch deeper than the baseline ≥ 3 scenarios. Tree shape comes from analysis/article-types.json → longHorizonRules, using scenarioCount for base scenarios plus wildcardCount (year-ahead) and scenarioBranchesPerScenario (election-cycle) where applicable:

Article type	Tree shape (minimum)	Leaves
quarter-ahead	4 base scenarios	4
year-ahead	4 base scenarios + 5 wildcards by default (recorded in wildcards-blackswans.md)	≥ 9
election-cycle	4 base scenarios × 3 governing-coalition branches	12

Probabilities sum to 100 % at every level (base set; per-scenario branch set). See the canonical scenario summary table and per-scenario branching Mermaid in analysis/templates/scenario-analysis.md. Year-ahead wildcards are authored in analysis/templates/wildcards-blackswans.md.

Counterfactual reasoning (mandatory at year/cycle)

devils-advocate.md MUST contain explicit **Counterfactual N — <name>** paragraphs at the minima below. Each counterfactual references at least one scenario ID from the sibling scenario-analysis.md (e.g. S2-Coalition-Collapse) and is anchored on a primary-source dok_id or URL. Gate LH-3 in 05-analysis-gate.md enforces the paragraph-count floor via grep.

Article type	Minimum counterfactual paragraphs
week-ahead / month-ahead	1
quarter-ahead	2
year-ahead	≥ 2 mandatory
election-cycle	≥ 3 mandatory

Paragraph structure (4–8 sentences): (a) restate the assumption being challenged, (b) construct the alternative world referencing the linked scenario ID, (c) declare the falsification trigger that would update the analysis.

Horizon stratification of strategic extensions

When horizonDays ≥ 90, every Family-C/D artifact in this methodology gains a horizon-class column (72h | week | month | quarter | year | cycle | election) on its probabilistic tables.

WEP terms (very likely, likely, roughly even, unlikely, very unlikely) carry an inline [horizon:<band>] tag within ± 80 characters of the term. WEP-band degradation rules — no very likely / very unlikely at year / cycle without ≥ 3 corroborated cycle-aged sources — are enforced at Pass-2 review per ext/long-horizon-forecasting.md §1.

Affected artifacts:

• scenario-analysis.md, devils-advocate.md, intelligence-assessment.md (Family C)
• risk-assessment.md, threat-analysis.md (Family A)
• cross-reference-map.md (Family B)
• forward-indicators.md, pestle-analysis.md (Family D — pestle-analysis.md blocking for year/cycle), wildcards-blackswans.md (Family D)

Single-horizon (horizonDays < 90) Family-C/D artifacts retain their existing column structure — no regression.

---

🛠️ Production Workflow — step-by-step

flowchart TD
    classDef trig fill:#1565C0,stroke:#0D47A1,color:#FFFFFF
    classDef step fill:#E8F5E9,stroke:#4CAF50,color:#1B5E20
    classDef gate fill:#FFF8E1,stroke:#FFC107,color:#3E2723
    classDef out fill:#F3E5F5,stroke:#7B1FA2,color:#311B92
    classDef vital fill:#FFF3E0,stroke:#FF9800,color:#BF360C,stroke-width:3px

    Trig[Family A + B complete<br/>→ Family C begins]:::trig

    S1[Step 1 — Branch paths<br/>→ scenario-analysis.md]:::step
    S2[Step 2 — International framing<br/>→ comparative-international.md]:::step
    S3[Step 3 — ACH + Red Team<br/>→ devils-advocate.md]:::step
    S4[Step 4 — Strategic bottom line<br/>→ intelligence-assessment.md]:::step
    S5[⭐ Step 5 — VITAL run-audit<br/>→ methodology-reflection.md]:::vital

    G{Gate — all 5 files<br/>produced + quality-gated?}:::gate

    O[Family C complete]:::out

    Trig --> S1
    Trig --> S2
    Trig --> S3
    S1 --> S4
    S2 --> S4
    S3 --> S4
    S4 --> S5
    S5 --> G
    G -->|pass| O
    G -->|fail| S1

---

✅ Family-C Completion Checklist

• scenario-analysis.md — ≥4 scenarios · probabilities sum to 100 % · branching Mermaid · early-warning indicators with ISO dates
• comparative-international.md — ≥5 peer countries · quantified outcomes · EU-law section · benchmark chart
• devils-advocate.md — n ≥ 3 hypotheses · evidence matrix ≥5 rows · fragility point · red-team hypothesis
• intelligence-assessment.md — 3–7 KJs · confidence labels · collection gaps · PIRs · tradecraft note
• methodology-reflection.md — evidence audit · confidence distribution · bias audit · forward adjustments
• All files use canonical color palette and 5-level confidence scale
• All files cross-link to Family A and Family B inputs

---

🔗 Template bindings

Template	Methodology section
analysis/templates/scenario-analysis.md	Part 1 above
analysis/templates/comparative-international.md	Part 2 above
analysis/templates/devils-advocate.md	Part 3 above
analysis/templates/intelligence-assessment.md	Part 4 above
analysis/templates/methodology-reflection.md	Part 5 above

---

📐 Cross-references to other methodology layers

• Upstream: synthesis-methodology.md (Family A) · structural-metadata-methodology.md (Family B)
• Downstream / parallel: electoral-domain-methodology.md (Family D — lens-specific extensions)
• Long-horizon authority (DRY): .github/prompts/ext/long-horizon-forecasting.md — scenario-tree depth, counterfactual minima, horizon stratification, IMF T+N stamps, cross-horizon citation
• Tier-C aggregation: .github/prompts/ext/tier-c-aggregation.md — multiplier table (1.7 / 2.0 / 2.5) and sibling-folder ingestion
• Frameworks: political-swot-framework.md · political-risk-methodology.md · political-threat-framework.md
• Master protocol: ai-driven-analysis-guide.md

---

🔐 ISMS Alignment

Control	How this methodology satisfies it
ISO 27001 A.5.7 (Threat intelligence)	Scenario + devils-advocate + intelligence-assessment constitute structured threat intelligence
ISO 27001 A.5.31 (Legal, regulatory)	Comparative-international maps EU-law intersection
NIST CSF ID.RA-3 (Threats identified)	ACH surfaces alternative threat interpretations
NIST CSF ID.RA-5 (Risks prioritised)	Scenario probabilities + impact enable risk ranking
CIS 17.5 (Incident response plan — decision support)	Intelligence-assessment maps to decision-support doctrine
GDPR Art. 35 DPIA methodology	Methodology-reflection provides audit trail

---

📜 Changelog

• v1.4 (2026-05-15) — Aligned version badge with header text (both now 1.4); bumped date to 2026-05-15.
• v1.3 (2026-04-25) — Explicit "Family C is exactly 5" close-out with placement rationale; methodology-reflection.md confirmed as 5th member and run-audit gate.

---

📄 Document Control

Owner: CEO (Intelligence Program) · Reviewer: CISO + Chief Analyst · Review Cycle: Quarterly Next Review: 2026-07-21 · Related: ai-driven-analysis-guide.md, synthesis-methodology.md, electoral-domain-methodology.md

---

Generated following Riksdagsmonitor Strategic Extensions Methodology v1.0 — Family C Depth Layer.