forward-indicators.md

May 3, 2026 · View on GitHub

🔭 Forward Indicators Template

📊 Actionable Watchlist of Dated Triggers That Will Update the Analysis
🎯 Leading Indicators · Threshold Rules · Confidence Labels · Ownership

📋 Document Owner: CEO | 📄 Version: 2.0 | 📅 Last Updated: 2026-05-01 (UTC) 🏢 Owner: Hack23 AB (Org.nr 559534-7807) | 🏷️ Classification: Public

📌 Template instructions: Produce on every run as a required deliverable, with at least the minimum indicator count for the article type's horizon band (see §Horizon Bands below). Save as analysis/daily/${ARTICLE_DATE}/${DOC_TYPE}/forward-indicators.md or, for review cycles, at analysis/weekly/${ISO_WEEK}/forward-indicators.md.

✨ What to produce: A single, authoritative watchlist — each indicator named, dated, threshold-defined, and tied to the downstream analysis it will update. Every indicator must be measurable from public MCP data.

📐 Template Contract — every fill of this template MUST satisfy this row.

Slot Value
Owning methodology per-artifact-methodologies.md
Owning gate check Check 8 (Family D — ≥ 10 dated indicators) — see 05-analysis-gate.md
Required inputs sibling Family A + scenario-analysis.md
Horizon band mixed (T+72h…T+90d) (per scripts/horizon-context.ts)
Output family Family D — Electoral & Domain
Aggregation order 7 of 30 in canonical order (see scripts/render-lib/aggregator/order.ts)
Reader Intelligence Guide row generated from forward-indicators.md (see scripts/render-lib/aggregator/reader-guide.ts)
Canonical evidence anchor | claim | evidence (dok_id / vote / MP intressent_id / primary-source URL) | retrieved_at | confidence | — every analytical claim row uses this schema.

Cross-reference: README.md §Template ↔ Methodology ↔ Gate-Check Matrix.

Slot	Value
Owning methodology	`per-artifact-methodologies.md`
Owning gate check	Check 8 (Family D — ≥ 10 dated indicators) — see `05-analysis-gate.md`
Required inputs	sibling Family A + scenario-analysis.md
Horizon band	mixed (T+72h…T+90d) (per `scripts/horizon-context.ts`)
Output family	Family D — Electoral & Domain
Aggregation order	7 of 30 in canonical order (see `scripts/render-lib/aggregator/order.ts`)
Reader Intelligence Guide	row generated from `forward-indicators.md` (see `scripts/render-lib/aggregator/reader-guide.ts`)
Canonical evidence anchor	`\| claim \| evidence (dok_id / vote / MP intressent_id / primary-source URL) \| retrieved_at \| confidence \|` — every analytical claim row uses this schema.

🧭 Horizon Bands

Registry-driven rendering: This template is a single file. The forwardIndicatorHorizons array in article-types.json drives which bands render for a given article type. Existing morning runs (week-ahead and shorter) produce the same output structure as before — back-compat is preserved.

Band Schema (conditional on `horizonDays`)

The forwardIndicatorHorizons array in article-types.json defines the exact band set per article type. The election band is always present as the terminal anchor. The pattern below shows how bands accumulate as the forecast horizon extends:

Condition	Active bands
`horizonDays = 0` (single-type, tier-c)	72h · week · month · election
`horizonDays = 7` (week-ahead)	72h · week · month · quarter · election
`horizonDays = 30` (month-ahead)	72h · week · month · quarter · year · election
`horizonDays = 90` (quarter-ahead)	week · month · quarter · year · election
`horizonDays = 365` (year-ahead)	month · quarter · year · cycle · election
`horizonDays >= 1460` (election-cycle)	quarter · year · cycle · election

WEP-Degradation Ladder (per-band ceiling)

As the forecast horizon extends, epistemic certainty decreases. The Words of Estimative Probability (WEP) ceiling degrades accordingly:

Band	Label	WEP ceiling	Rationale
T+72h	72 hours	very likely / very unlikely	Near-term; high confidence from scheduled events
T+7d	week	likely / unlikely	Short-term; most parliamentary triggers known
T+30d	month	likely / unlikely	Medium-term; some uncertainty in scheduling
T+90d	quarter	roughly even / about even	Extended; multiple intervening variables
T+365d	year	roughly even (mandatory unless ≥3 corroborated sources)	Long-range; structural uncertainty dominates
T+1460d	cycle	roughly even / unlikely / very unlikely; never likely/very likely without ≥3 cycle-aged sources	Ultra-long; scenario-driven only

Minimum Indicator Counts (per article type)

Article type	Min indicators	Bands
Week-ahead (`horizonDays=7`)	6	72h, week, month, quarter, election
Month-ahead (`horizonDays=30`)	8	72h, week, month, quarter, year, election
Quarter-ahead (`horizonDays=90`)	10	week, month, quarter, year, election
Year-ahead (`horizonDays=365`)	12	month, quarter, year, cycle, election
Election-cycle (`horizonDays>=1460`)	15	quarter, year, cycle, election

Note: Single-type and tier-c-aggregation articles (morning runs, evening analysis) use the 4-band schema (72h / week / month / election) with a floor of 6 indicators. The election band is always present as the terminal anchor across all article types.

🔄 Tradecraft Context

Field	Value
F3EAD stage	`Analyze / Disseminate` (forward-looking watchlist drives next collection cycle)
PIRs	`every indicator ties to one or more standing PIRs — name them per indicator`
Admiralty floor	`A1 for measurable MCP-sourced thresholds; B2 acceptable for derived composite indicators`
SATs used	`Indicators & Signposts; Key Assumptions Check; Outside-In Thinking`
ICD 203 standards applied	`uncertainty, alternative analysis, consistency/change, customer relevance`

See political-style-guide.md for canonical F3EAD / PIR catalog / Admiralty Code / ICD 203 / WEP / SATs definitions.

📋 Watchlist Context

Field	Value
Watchlist ID	`FWI-YYYY-MM-DD-NNN`
Generated	`YYYY-MM-DD HH:MM UTC`
Scope	`e.g., week 17 / month 2026-04 / run-specific`
Indicator count	`N`
Overall Confidence	`🟩 HIGH`

🧭 Indicator Dashboard

⚠️ Illustrative example below — the Gantt milestones, indicator IDs, thresholds, and deadlines are drawn from a worked 2026 scenario. Replace every milestone, date, indicator, threshold, and owner with run-specific values before publishing. Do not treat this example as an actual monitoring commitment.

gantt
    title Forward-indicator timeline
    dateFormat YYYY-MM-DD
    section Parliamentary
    FiU48 chamber vote (FI-01)         :milestone, 2026-04-24, 0d
    CU13 chamber vote (FI-02)          :milestone, 2026-04-29, 0d
    FöU proposition tabling (FI-03)    :milestone, 2026-05-12, 0d
    section Polling
    Q2 SIFO release (FI-04)            :milestone, 2026-06-30, 0d
    Novus mid-month (FI-05)            :milestone, 2026-07-15, 0d
    section Macro
    SCB PR0101 May (FI-06)             :milestone, 2026-06-12, 0d
    IMF WEO update (FI-07)             :milestone, 2026-07-22, 0d
    section External
    EU Commission response (FI-08)     :milestone, 2026-05-15, 0d

🗂️ Indicator Register

#	Indicator	Measurable via	Trigger threshold	Deadline	Owner (analysis file)	Confidence
FI-01	FiU48 chamber vote outcome	`search_voteringar(bet=FiU48)`	Coalition unanimity or ≥ 1 `Avstår`	2026-04-24	`risk-assessment.md` R1	🟦 VERY HIGH
FI-02	CU13 chamber vote outcome	`search_voteringar(bet=CU13)`	Any close result (< 10-vote margin)	2026-04-29	`risk-assessment.md` R3	🟩 HIGH
FI-03	FöU defence proposition tabling	`get_propositioner`	Tabled or delayed	2026-05-12	`threat-analysis.md` T2	🟩 HIGH
FI-04	Q2 2026 SIFO release	Published by Kantar Sifo	Government-bloc gap ≥ −6 pp	2026-06-30	`election-2026-analysis.md`	🟩 HIGH
FI-05	Novus mid-July	Published by Novus	Consistency with FI-04 direction	2026-07-15	`scenario-analysis.md`	🟩 HIGH
FI-06	SCB PR0101 May CPI	`scb` query_table PR0101	Pump-price index ↓ ≥ 3 %	2026-06-12	`voter-segmentation.md` S1	🟦 VERY HIGH
FI-07	IMF WEO July update	`tsx scripts/imf-fetch.ts`	Sweden GDP revision	2026-07-22	`comparative-international.md`	🟩 HIGH
FI-08	EU Commission communication on fuel-tax cut	Commission press room	Any formal letter / opening of review	2026-05-15	`risk-assessment.md` R2	🟧 MEDIUM
FI-09	SD public posture on fuel-tax sunset	`search_anforanden(parti=SD)`	Any ambiguity signal	Rolling	`coalition-mathematics.md`	🟧 MEDIUM
FI-10	Skatteverket guidance publication	Skatteverket site	Publication date	2026-05-30	`implementation-feasibility.md` D2	🟩 HIGH

🧪 Indicator Detail — Example (repeat per FI)

FI-01 — FiU48 chamber vote outcome

Attribute	Value
Monitoring method	MCP call `search_voteringar(bet=FiU48, rm=2025/26)`
What a "positive" signal means	Unanimous government-bloc Ja → confirms coalition discipline claim
What a "negative" signal means	Any coalition MP `Avstår` or `Nej` → coalition fracture signal
Who changes?	`risk-assessment.md` R1 score; `devils-advocate.md` ACH hypothesis H3; `synthesis-summary.md` §Finding 1
Fallback	If vote delayed, re-check 2026-04-25, 2026-04-28
Confidence	🟦 VERY HIGH — Riksdag voting records are primary-source public data

🔁 Update Rules

Rule	When it fires
Auto-rerun	Any indicator reaches its deadline
Re-score	Indicator threshold is breached
Promote	Indicator exceeds threshold for 2 consecutive observations → risk register upgrade
Retire	Indicator passes deadline without breach → archive with observation note

📅 This-Week Watch Window

Day	Indicator	Action
2026-04-22	FI-01 prep	Pre-load `search_voteringar` query
2026-04-24	FI-01	Record outcome, rerun dependent analyses
2026-04-25	FI-09	Check SD spokesperson statements
2026-04-29	FI-02	Record outcome

🧭 Cross-File Impact Map

graph LR
    FI01["🗓️ FI-01<br/>FiU48 vote"] --> RA["⚠️ risk-assessment.md"]
    FI01 --> DA["😈 devils-advocate.md"]
    FI01 --> SYN["🧩 synthesis-summary.md"]
    FI04["🗓️ FI-04<br/>Q2 SIFO"] --> EL["🗳️ election-2026-analysis.md"]
    FI04 --> SCN["🔮 scenario-analysis.md"]
    FI06["🗓️ FI-06<br/>SCB PR0101"] --> VS["🫂 voter-segmentation.md"]
    FI08["🗓️ FI-08<br/>EU Commission"] --> CI["🌍 comparative-international.md"]

    style FI01 fill:#FF9800,color:#FFFFFF
    style FI04 fill:#7B1FA2,color:#FFFFFF
    style FI06 fill:#1565C0,color:#FFFFFF
    style FI08 fill:#C2185B,color:#FFFFFF
    style RA fill:#D32F2F,color:#FFFFFF
    style DA fill:#FFC107,color:#000000
    style SYN fill:#4CAF50,color:#FFFFFF
    style EL fill:#4CAF50,color:#FFFFFF
    style SCN fill:#7B1FA2,color:#FFFFFF
    style VS fill:#4CAF50,color:#FFFFFF
    style CI fill:#1565C0,color:#FFFFFF

📎 Sources

Source	Use
`riksdag-regering` MCP	Voteringar, anföranden, dokument, kalenderhändelser
`scb` MCP	CPI, pump-price index, labour statistics
`imf` (scripted)	Macro-fiscal updates
`world-bank` MCP	Governance / comparator indicators
Kantar Sifo, Novus, Demoskop	Public polling
EU Commission press room	State-aid communications

Document Control

Template path: /analysis/templates/forward-indicators.md
Referenced by: ai-driven-analysis-guide.md § Step 5
Classification: Public
Next Review: 2026-07-21

✅ Pass-2 Self-Audit Checklist (v4.4 — required)

Purpose: AI-FIRST principle requires a Pass-2 read-back-and-improve. After producing this artifact in Pass 1, re-read it end-to-end and verify each item below. Document any remediation in methodology-reflection.md §"Pass-2 audit log". Any unchecked ❌ box at the end of Pass 2 forces a Pass-3 rewrite of the affected section.