🔄 Cross-Run Diff Template

May 3, 2026 · View on GitHub

📌 Template Instructions — Copy to analysis/daily/$ARTICLE_DATE/$SUBFOLDER/cross-run-diff.md. Bayesian delta vs. the previous run of the same article type: what changed in data and assessment. See per-artifact-methodologies.md §cross-run-diff. N/A when this is the first ever run of its type.

🎯 Purpose — Track evolution of political assessment across runs. For each prior-run finding, apply Bayesian update based on new evidence to produce a posterior assessment with an explicit WEP band shift.

🔄 Tradecraft Context

Artifact role — Compares the current run against the immediately previous run of the same article type to identify material changes in evidence, assessment, and confidence.
When to use — Required for every non-initial run. If this is the first ever run for the article type, mark this artifact N/A and state that no prior baseline exists.
Primary inputs — Current-run core artifacts, prior-run core artifacts, prior synthesis-summary.md, prior reference-analysis-quality.md, and any newly surfaced source evidence that materially changes the posterior judgement.
Analytical standard — Use explicit Bayesian-style updating: identify what changed, explain whether each change strengthens, weakens, or leaves unchanged the prior assessment, and record any resulting WEP band movement with concise justification.
Output requirement — Focus only on deltas that matter for editorial judgement, risk framing, or confidence. Do not restate unchanged background unless needed to explain a shift.

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

Slot Value
Owning methodology per-artifact-methodologies.md
Owning gate check Supplementary Check 11 (gate-required when ANALYSIS_RUN_COUNT ≥ 2) — see 05-analysis-gate.md
Required inputs previous run folder of same article type
Horizon band per-run (per scripts/horizon-context.ts)
Output family Operational Supplementary
Aggregation order appended (alphabetical, after canonical block) (see scripts/render-lib/aggregator/order.ts)
Reader Intelligence Guide row generated from cross-run-diff.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	Supplementary Check 11 (gate-required when ANALYSIS_RUN_COUNT ≥ 2) — see `05-analysis-gate.md`
Required inputs	previous run folder of same article type
Horizon band	per-run (per `scripts/horizon-context.ts`)
Output family	Operational Supplementary
Aggregation order	appended (alphabetical, after canonical block) (see `scripts/render-lib/aggregator/order.ts`)
Reader Intelligence Guide	row generated from `cross-run-diff.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.

📋 Document Metadata

Field	Value
Report ID	`[REQUIRED: XRD-YYYY-MM-DD-runNN]`
Current Run	`[REQUIRED: {type}-run{N}, YYYY-MM-DD]`
Prior Run	`[REQUIRED: {type}-run{N-1}, YYYY-MM-DD]`
Days Between Runs	`[REQUIRED: #]`
Confidence in Delta	`[REQUIRED: 🟢 / 🟡 / 🔴]`

1️⃣ Prior-Run Header

Prior run directory — [REQUIRED: analysis/daily/YYYY-MM-DD/{type}-run{N-1}/]
Prior run date — [REQUIRED]
Prior run gate status — [REQUIRED: ✅ PASSED / ❌ FAILED]

Key findings from prior synthesis-summary.md:

[REQUIRED: 1-line prior finding + its WEP band]
[REQUIRED]
[REQUIRED]
[REQUIRED]
[REQUIRED]

Prior scores (carry forward from prior reference-analysis-quality.md §Overall judgement):

Metric	Prior value
Weighted quality score	`[#/10]`
Reliability score (MCP)	`[#]`
Coverage (core artifacts)	`[#]/23`
P0 claims	`[#]`
Admiralty A-rated sources	`[#]`

2️⃣ New Evidence Since Prior Run

Evidence ID	Source	Date	Artifact affected	Admiralty
`[dok_id / vote ID / URL]`	`[riksdag-regering / scb / IMF / press]`	`[YYYY-MM-DD]`	`[current run artifact]`	`[A1/…]`

Minimum — ≥ 3 new evidence rows when days-between ≥ 3; ≥ 1 otherwise. Zero new evidence on a multi-day gap → flag prior run's stability claim.

2.1 IMF Vintage Delta

Indicator	Prior vintage	Current vintage	Prior value (year)	Current value (year)	Δ (pp)	Action
`WEO:NGDP_RPCH`	`WEO-2025-10`	`WEO-2026-04`	`[e.g. 1.8 % (2026)]`	`[e.g. 2.1 % (2026)]`	`+0.3`	`[update / no-op]`
`WEO:GGXWDG_NGDP`	`WEO-2025-10`	`WEO-2026-04`	`[e.g. 32.1 % (2027)]`	`[e.g. 32.4 % (2027)]`	`+0.3`	`[update]`
`WEO:PCPIPCH`	`WEO-2025-10`	`WEO-2026-04`	—	—	—	—

Rule: any headline indicator whose projection shifted by > 0.5 pp across vintages gets a [VINTAGE-DELTA] annotation in the article commentary, and prior-run citations are rewritten to the current vintage. Add a sentence in §3 Bayesian Update explaining the methodology shift. Stale-vintage citations (> 6 months old per .github/aw/ECONOMIC_DATA_CONTRACT.md § Vintage staleness rule) MUST be refreshed before publish.

3️⃣ Bayesian Update — Key Judgments

For each prior KJ, apply posterior ∝ prior × likelihood(new evidence | KJ) and state the new WEP band.

KJ 1 — `[REQUIRED: prior KJ headline]`

Prior WEP — [e.g. Likely (55-70%), 6-month horizon]
New evidence — [list evidence IDs from §2 that bear on this KJ]
Likelihood assessment — [does new evidence confirm / disconfirm / condition?]
Posterior WEP — [e.g. Very Likely (70-85%), 6-month horizon]
Shift rationale — [2-3 sentences naming the evidence and its Admiralty]

(Repeat for each prior KJ.)

4️⃣ Scenario Probability Shifts

Scenario (from prior `scenario-analysis.md`)	Prior P	Posterior P	Δ	Trigger evidence
`[REQUIRED]`	`[%]`	`[%]`	`[±%]`	`[evidence ID]`
`[REQUIRED]`	`[%]`	`[%]`	`[±%]`	`[evidence ID]`
`[REQUIRED]`	`[%]`	`[%]`	`[±%]`	`[evidence ID]`

Sum of posteriors must equal 100%.

5️⃣ Risk & Threat Register Deltas

Prior row	Prior L×I	Posterior L×I	Δ	New row?
`[REQUIRED]`	`[score]`	`[score]`	`[±]`	`[n]`

Rows added this run: [#]. Rows retired: [#] (with justification).

6️⃣ Stakeholder Position Changes

Actor	Prior position	Posterior position	Evidence
`[REQUIRED: MP + intressent_id / party / myndighet]`	`[support/oppose/ambiguous]`	`[…]`	`[speech ID / vote ID]`

7️⃣ Framing Evolution

Cross-reference the longitudinal frame record in media-framing-analysis.md:

Prior dominant frame — [REQUIRED]
Current dominant frame — [REQUIRED]
Shift driver — [data / political / external event]
Counter-frame status — [emerging / stable / absorbed]

8️⃣ Methodology Drift Check

Any methodology or template version changes between the two runs? List them and their impact on comparability:

Change	Prior version	Current version	Comparability impact
`[REQUIRED]`	`[vX.Y]`	`[vX.Y]`	`[none / rescored / incomparable]`

If comparability is incomparable, downgrade overall diff confidence to 🔴 and flag in methodology-reflection.md.

9️⃣ Summary Table

Dimension	Prior	Current	Δ	Direction
Quality score	`[#/10]`	`[#/10]`	`[±]`	`[↑/↓/=]`
KJ count	`[#]`	`[#]`	`[±]`	`[…]`
P0 claims	`[#]`	`[#]`	`[±]`	`[…]`
Admiralty A-rated	`[#]`	`[#]`	`[±]`	`[…]`
Scenario entropy	`[#]`	`[#]`	`[±]`	`[…]`
Top risk L×I	`[#]`	`[#]`	`[±]`	`[…]`

One paragraph (3–5 sentences) summarising the single largest shift in assessment, with WEP band + horizon + Admiralty of the triggering evidence. This paragraph feeds directly into the next synthesis-summary.md BLUF.

🔗 Cross-References

Methodology: ../methodologies/per-artifact-methodologies.md#cross-run-diff
Prior synthesis: [REQUIRED: link to prior run's synthesis-summary.md]
Current synthesis: synthesis-summary.md
SAT canon: ../methodologies/osint-tradecraft-standards.md

Template version: v1.2 · Last updated: 2026-04-25

✅ 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.