World Factbook 2025 Context Integration and Data Quality Report

Status

Current implementation report for IARPG-OPS-2 2.0.19-wip. The imported dataset is published as an archival context snapshot, not as a live country feed, current geopolitical assessment, legal authority, development ranking, or intelligence product.

Purpose

Document how the supplied Factbook2025all_countries.csv was preserved, audited, normalized, bounded, published, connected to the international-coverage roadmap, distributed through public site content, and linked into active .uai memory without converting descriptive country context into alignment, morality, threat, loyalty, dangerousness, or recruitability scores.

Scope

This report covers:

• byte-preserving source intake and provenance;
• CSV structure, entity identity, category inventory, and duplicate-code analysis;
• the six-lens public context model;
• mapping to the 195-country international-coverage roadmap;
• public page, JSON, schema, profile, search, answer, glossary, and editorial integration;
• international-fairness and archival-truth boundaries;
• memory, validation, and release implications.

It does not independently verify every factual statement in the 32,594-row source, certify the source as current, or establish contemporary legal, political, economic, demographic, environmental, or institutional truth for any jurisdiction.

Executive Summary

The project owner supplied a 7,471,330-byte CSV containing 32,594 data rows, 260 named entities, 247 distinct code values, and 13 source categories. The file is preserved byte-for-byte at docs/source-files/world-factbook-2025/Factbook2025all_countries.csv with SHA-256:

f98d6a038a9b041ea208b1d463f5b1dedc28a5328d19cd030cdd1c276997c9e9

The source header labels one column ISO2, but that field cannot safely serve as a unique entity key. Eight code values are reused across multiple records, including GB for the United Kingdom, Guernsey, Jersey, and the Isle of Man; AU for Australia and two external areas; and PS for the West Bank and Gaza Strip. The publication therefore retains the imported value as sourcecode and creates an independent stable entityid from the entity name. No entity is silently merged because of a colliding code.

A curated public publication contains 260 profiles and 11,781 facts organized through six equal lenses: place, people, institutions, economy, infrastructure, and environment. It maps source context to 195 of 195 countries in the site’s international-coverage roadmap and retains 64 additional source entities—including territories, dependencies, oceans, special areas, and a supranational record—as visibly different entity classes.

The public explorer deliberately omits Military and Security, Terrorism, and Transnational Issues from its synthesized profile layer. Those rows remain in the preserved source but are not republished as public scoring or operational content. The explorer calculates no composite score, rank, alignment, risk, morality, loyalty, danger, recruitability, or targetability measure.

The CIA announced on February 4, 2026 that The World Factbook had sunset. The site therefore describes this material as a 2025 archival snapshot and preserves per-field estimate years where detectable. Contemporary statements require current primary, official, locally authored, or local-language evidence and an explicit verification date.

Evidence Reviewed

Supplied source

• Factbook2025allcountries.csv
• CSV header: Country, ISO2, Category, Field Name, Content, Value
• Source SHA-256 and byte count recorded above

Generated publications

• world-context-2025.json
• world-context-2025-index.json
• world-context-2025-audit.json
• world-context-2025.schema.json
• world-context-2025-index.schema.json
• world-context-profile.schema.json
• Per-profile records under world-context/entities/

Repository implementation

• scripts/buildworldcontext.py
• world-context.php
• assets/js/world-context.js
• assets/css/site.css
• app/layout.php
• app/content.php
• global.php
• world.php
• start.php
• index.php
• editorial.php
• glossary.php
• docs/index.php

Current authority and methodological context

• CIA: “Spotlighting The World Factbook as We Bid a Fond Farewell”
• International Fairness Methodology
• International Coverage Roadmap
• Editorial and Evidence Method
• [Global Alliance and Alignment Atlas](global-alliance-and-alignment-atlas.md#reviewed-synthesis)
• [Alliance, Faction, and Network World-System Integration Report](alliance-faction-and-network-world-system-integration-report.md#findings)
• [Planetary Systems Research Integration and Scenario Modeling Report](planetary-systems-research-integration-and-scenario-modeling-report.md#findings)

Findings

1. The source is broad enough to improve context, but not current enough to stand alone

The CSV provides repeatable descriptive structure across geography, people and society, government, economy, communications, energy, transportation, environment, and other categories. This makes it useful for orientation and fictional worldbuilding. It does not make the material a current or exhaustive account of any place.

The collection year and the measurement year are different concepts. A source preserved as a 2025 snapshot may contain a population estimate from one year, a GDP estimate from another, and legal or constitutional prose that changed after collection. The public records therefore carry estimate_year beside individual facts when a year can be detected and retain a separate snapshot-year boundary.

2. The imported code column is not a unique identifier

The audit found eight reused code values:

A key derived directly from that field would overwrite or conflate distinct records. The publication resolves this by:

1. preserving the imported value as source_code;
2. assigning a stable, human-readable entity_id derived from the entity name;
3. recording every collision in the machine-readable audit;
4. classifying entities separately from the 195-country roadmap;
5. treating the State of Palestine roadmap entry as a source-context composition of the West Bank and Gaza Strip records rather than silently inventing a third source row.

3. Equal structure improves comparability without implying equivalence

The public publication uses six lenses:

1. Place and physical context — location, terrain, climate, scale, hazards, resources, and settlement distribution.
2. People and settlement — population, languages, urbanization, age, education, and health context.
3. Institutions and public structure — capital, government form, legal system, administrative divisions, constitution, and international participation.
4. Economy and livelihoods — overview, production, growth, prices, employment, industries, trade, and agriculture.
5. Connectivity and infrastructure — internet, electricity, energy, airports, ports, railways, and merchant transport.
6. Environment and resource pressure — environmental issues, emissions, water, withdrawal, and international agreements.

Every profile uses the same presentation model, but missing data remains missing and no composite score is calculated. Equal structure means equivalent questions and visible evidence boundaries—not identical depth, identical status, or identical conclusions.

4. All 195 roadmap countries have an archival context route

The alias and composition mapping resolves every country in the international roadmap. It handles naming differences such as Burma (Myanmar), Congo, Democratic Republic of the, Cote d'Ivoire, Gambia, The, Korea, North, Korea, South, Sao Tome and Principe, and Turkiye.

The remaining 64 source entities stay available as additional context but are not relabeled as sovereign countries. This prevents oceans, dependencies, special areas, and the European Union from being flattened into the same political class.

5. Context is not alignment, intent, legitimacy, or risk

Geography, demographics, language, religion, economy, infrastructure, environment, government form, disability, diagnosis, or identity do not establish:

• who a country or institution aligns with;
• whether a relationship is voluntary, coerced, transactional, formal, or disputed;
• whether an actor is legitimate or illegitimate;
• whether a population is dangerous or recruitable;
• whether conduct is lawful, ethical, hostile, friendly, stable, or inevitable.

The World Context Explorer is therefore upstream of the Alliance Observatory and Global Intelligence Atlas. It helps readers ask better questions; current relationship claims require separate evidence, confidence, disagreement, and correction records.

6. The public layer excludes high-risk source categories

The source contains 13 categories, including Military and Security, Terrorism, and Transnational Issues. The public profile synthesis excludes those three categories. This avoids turning archival military, terrorism, trafficking, illicit-drug, or other transnational prose into a targeting layer, threat score, or unreviewed operational profile.

The preserved CSV remains available for provenance and audit. Exclusion from public synthesis is not deletion from the source record.

7. The browser architecture avoids a multi-megabyte initial download

The full curated dataset is approximately the size of the source and remains directly downloadable. The interactive page initially requests a compact index and loads one profile JSON file on demand. This keeps the initial interface responsive while preserving the complete machine publication.

Search, filtering, comparison, and selected-profile state operate in the browser. The implementation uses no third-party runtime, analytics endpoint, account, telemetry service, local storage, session storage, or cross-origin data request.

8. Public content now explains the world before presenting its relationships

The site routes readers from Start Here and World through World Context before Alliance Observatory, Faction Directory, Campaign State, Network Influence, and Scenario Modeling. The Global Intelligence Atlas now makes explicit that place-level context and institution-level alignment are different layers.

Direct answers and glossary entries explain the archival boundary, entity identity, estimate-year distinction, and no-score rule. The editorial method defines how future agents must update or supersede context.

Decisions or Recommendations

1. Keep the source immutable. Preserve the uploaded CSV byte-for-byte and record its SHA-256 in the source manifest.
2. Keep source_code descriptive. Never promote the imported ISO2 column to a unique or normative identifier without a separately verified crosswalk.
3. Use stable entity_id values. Entity IDs must remain stable across regenerations even if display labels or aliases change.
4. Retain entity classes. Countries, territories, dependencies, oceans, special areas, and supranational records must remain distinguishable.
5. Preserve per-field years. Never present the snapshot year as though it were the estimate year of every field.
6. Require current evidence for current claims. Any public sentence using “current,” “now,” “today,” “as of,” or equivalent language must be reverified against primary, official, locally authored, or local-language sources.
7. Do not infer relationships from context. Route alignment analysis through relationship evidence and confidence records rather than demographic or economic proxies.
8. Do not create country scores. Prohibit composite development, stability, threat, loyalty, danger, recruitability, or moral rankings.
9. Show disagreement and missingness. Missing values, code collisions, contested political status, and source limitations must stay visible.
10. Add claim-level freshness metadata in a future round. A later country-profile composer should attach verified_utc, source jurisdiction, source language, authority class, and correction history to every contemporary claim.
11. Keep the interactive publication browser-local. Continue to use the compact index plus on-demand profile records and avoid analytics or identity collection.
12. Run a comparative fairness audit before dedicated country prose is promoted. Review adjective intensity, rights coverage, affected-community visibility, source diversity, local terminology, uncertainty, and gameplay consequences across regions.

Risks and Limitations

• The source was supplied as a finished CSV; this release does not establish the original extraction methodology, completeness, licensing chain, or transformation history before upload.
• The dataset contains historical and time-sensitive claims. Some fields may already have been stale before the 2025 snapshot was assembled.
• Year extraction is heuristic and may identify a referenced year that is not the principal estimate year in unusually structured text.
• Entity classification is an editorial aid, not a legal determination of sovereignty or international recognition.
• The 195-country mapping is a site-navigation crosswalk, not a claim that all political-status questions are resolved.
• A complete profile can still be unbalanced if the source itself is uneven, centrally authored, outdated, or missing local perspectives.
• The public exclusion rules reduce harmful reuse but do not independently verify the remaining 11,781 facts.
• Browser-local comparison can encourage false equivalence if readers ignore units, years, methodologies, and missing values; the interface therefore avoids rankings and includes repeated boundary notices.
• The source publisher’s sunset means there will be no future Factbook updates to rely on as a continuing series. A replacement workflow must use multiple current sources rather than treating this archive as self-updating.

Validation Performed

The 2.0.19-wip validation suite checks, at minimum:

• exact source filename, byte count, and SHA-256;
• CSV header and 32,594 data rows;
• 260 unique entity names and 247 unique source-code values;
• exactly eight recorded code-collision groups;
• 260 full profiles and 260 per-profile JSON files;
• 11,781 curated public facts;
• six declared public lenses;
• 195 of 195 roadmap-country mappings;
• no synthesized public fact from Military and Security, Terrorism, or Transnational Issues;
• schema validation for the full publication, compact index, and every profile;
• stable profile URLs and entityid values that do not rely on sourcecode uniqueness;
• page, answer, glossary, editorial, source-manifest, report, pointer-ledger, and UAI links;
• browser-local search, profile loading, comparison, mobile containment, and no external request or persistent-storage use;
• static syntax, machine records, memory links and anchors, source integrity, accessibility structure, delivery budgets, sitemap metadata, package integrity, and fresh-extraction equality.

Detailed results are published in the release validation and focused World Context evidence files.

Memory References

• World context data hot memory
• International coverage
• Regional intelligence atlas
• International fairness
• Source provenance
• Source map
• Research index
• Architecture
• Progress
• Test plan
• Long-term memory pointer ledger

Related Durable Documents

• [International institutional fairness and evidence audit](equal-standards-across-unequal-records-international-institutional-fairness-rights-and-evidence-audit.md#evidence-asymmetry-analysis)
• [Global Alliance and Alignment Atlas](global-alliance-and-alignment-atlas.md#reviewed-synthesis)
• [How Alliances Actually Work](how-alliances-actually-work.md#reviewed-synthesis)
• [Informal Alignments, Patrons, Proxies, Intermediaries, and Issue-Specific Coalitions](informal-alignments-patrons-proxies-intermediaries-and-issue-specific-coalitions.md#reviewed-synthesis)
• [Alliance, Faction, and Network World-System Integration Report](alliance-faction-and-network-world-system-integration-report.md#findings)
• [Planetary Systems Research Integration and Scenario Modeling Report](planetary-systems-research-integration-and-scenario-modeling-report.md#findings)
• [UAI Memory Organization Report](uai-memory-organization-report.md#findings)

Supersession Status

Current for IARPG-OPS-2 2.0.19-wip. A later report may supersede the implementation details when a multi-source, claim-level-current country context system replaces or materially extends this archival snapshot. The preserved CSV and its provenance record must remain available even after supersession.