ACE World Intelligence

Governed public-issue sensemaking on the shared ACE foundation.

ACE World Intelligence is an independently versioned ACE domain product. Its installable package is an inert, JSON-only Domain Pack for making sense of a changing public issue. It separates admitted records, attributed claims, corroboration, disputes, ACE inference, unknowns, and conditional scenarios instead of presenting one opaque summary as truth.

It supplies World vocabulary, source mappings, material-change policy, personas, synthesis policy, connectors, fixtures, and product evidence. It does not implement a second reasoning runtime, graph, state store, authority system, detector engine, or feedback loop.

Install · Architecture · Proof · Roadmap · Security · Contributing

• Distribution: ace-domain-world-intelligence 0.10.0
• Requires: Python 3.12 and ace-core>=0.6.0,<0.7
• Artifact boundary: JSON-only, data-only, inert
• Release identity: v0.10.0 and 0.10.0 on PyPI. Public-artifact closeout is recorded separately from the immutable source tag.

What you install, and what you get

The product is split into three layers, and this repository owns only the third.

Installing the Domain Pack adds data, not behaviour. The wheel contains no .py, no entry points, and no install hooks; nothing in it executes on install or import. ACE Core compiles those JSON modules and does the reasoning. If you want to sense a live official source, you additionally install a separately reviewed connector — see Connector boundary.

Install

Install 0.10.0 on Python 3.12 with either command below; a public clean install resolves ace-domain-world-intelligence==0.10.0 and ace-core==0.6.0.

With uv:

uv add "ace-domain-world-intelligence==0.10.0"

With pip:

pip install "ace-domain-world-intelligence==0.10.0"

Either command also brings in ace-core>=0.6.0,<0.7. It does not bring in the Federal Register connector; that is a deliberate boundary, not an omission.

Resolve the pack data from the installed distribution:

import json
from importlib.resources import files

manifest = json.loads(files("domain_packs.world_intelligence").joinpath("manifest.json").read_text(encoding="utf-8"))
print(manifest["metadata"]["pack_id"])  # world_intelligence

Develop from a source checkout

To develop the pack, or to run the acceptance harnesses and the P2C connector packet, work from a checkout alongside an ACE Core checkout or installation:

# from the root of this checkout
export REPO="$PWD"
export ACE=/path/to/ace-core
export PYTHONPATH="$REPO:$REPO/adapters/federal_register_source/src:$ACE"
export PY="$ACE/.venv/bin/python"

PYTHONPATH above includes the connector source so that the P2C packet runs. Drop that middle entry and the P2C modules skip rather than fail.

Product loop

authorized evidence → Observation → Entity Snapshot
                           ├────────→ Signal ──┐
                           └────────→ Shift ───┼→ Case / Brief → Decision → Outcome
                                               └→ governed feedback proposal

This is a typed DAG, not a forced pipeline. A Shift need not become a Signal, a Signal need not become a Brief, and no downstream resource grants itself authority.

Domain scope

World Intelligence is intended to cover public events, actors, institutions, issues, policies, claims, commitments, actions, reactions, corrections, and source relationships. Its user-facing products can include Reality Briefs, event summaries, standing investigations, material-change alerts, and personal briefings. These are World-domain types and policies; shared graph, detection, routing, synthesis, authority, and feedback machinery remains in ACE.

Current milestone

P2A proved that a materially different, JSON-only World Intelligence pack compiles through the unchanged ACE pack compiler (pack_ir:683de57a71669814e507d07d65a109db, seven conformance tests).

P2B freezes the golden 72-hour public-issue scenario: the synthetic, hermetic, redistributable meridia_reservoir_release_72h packet. It pins eight timestamped source records across four provenance families (including two syndicated copies and a publisher correction), a nineteen-entity graph exercising all eleven relation types, an official status change, four material categorical shifts plus one numeric shift, one runtime-routable signal, and an exact Reality Brief whose seventeen statements carry all seven epistemic statuses. Nine negative vectors fail closed, from missing attribution through divergent replay.

The additive P2B PREPARED replay now compiles and activation-binds detection v1alpha2, admits all eight source records as PREPARED Observations, reproduces all five Shifts, emits only the intended four Signals, and resolves all four persona routes. The claim-corroboration Shift is represented through ACE's generic resource-set admission without inventing a Signal. The frozen scenario and default 0.1.0 pack identities remain unchanged.

The replay also freezes those developments into ACE's public immutable Case contract. Four routed Signals plus the non-routed corroboration Shift form one five-member Case that transitively binds the complete 28-resource scenario closure. World supplies the case type and purpose; ACE owns the identity, temporal validation, lineage resolution, admission, and replay.

An additive Case-bound packet now consumes ACE's public CaseBriefSynthesisService over that exact Case. It admits the routed derivations through ACE's durable PREPARED ledger without duplicating any source record, derives one compatible template and persona scope from the four exact attention receipts, freezes all 26 closure resources as Core context, and emits one governed, grounded, atomically persisted, deterministically replayable Reality Brief that carries the Case in lineage: brief:8fb3173069eca502652b1c9c004c92e6. WI-CR-005 is closed.

Per-statement epistemic status (WI-CR-002) and derivation-family independence (WI-CR-003) are now closed through public, domain-neutral ACE contracts. World declares its seven-label vocabulary; ACE validates each statement against its exact supports and proves that corroborated spans at least two admitted derivation families. The Ledger report and the independent Basin gauge satisfy that rule, while Coastal Wire and Harborview collapse into the Ledger root despite being different publishers. The durable projection discloses the exact root-to-member partition used by the proof. Supersession-impact enumeration (WI-CR-004) is now closed too, so no public-platform contract request remains open. When a record is superseded, ACE enumerates exactly what depended on it, directly and transitively, and appends that answer as an additive immutable record. Both frozen corrections are exercised: the Ledger correction reaches 11 resources and 9 Brief claims, the Order supersession reaches 7 resources and 5 claims, and each projection discloses the resources it found unaffected so the boundary is visible rather than inferred. Impact means dependency, not falsehood — the prior Brief keeps its exact identity, replays byte-identically, and is never rewritten.

P2C adds the first official-source sensing proof without mutating that frozen pack. A separate, JSON-only activation pack (pack_ir:1847032fc5301bba9b6f85d3d091400d) maps one exact Federal Register document, while a separately versioned, transport-injected adapter validates the exact source, artifact, URI, network attestations, and closed payload. ACE then admits one acquisition receipt, source snapshot, visibly LIVE Observation, LIVE Entity Snapshot, and admission receipt as one atomic transaction. Exact replay and fresh-service restart replay perform no second capture.

The adapter's additive 0.2.0 network boundary is now implemented and opt-in. On 2026-08-10 it completed one credential-free, redirect-free, DNS-pinned retrieval from the exact reviewed API URI: HTTP 200 application/json, 4,227 characters, body SHA-256 62e5f5da8f712542926093e897366cb49916afd92c4233e12d2a791581cf6571. That is acquisition evidence only. The deterministic governed P2C admission remains recorded-transport conformance; the external capture has not been promoted to a Signal, Shift, Case, Brief, or legal conclusion.

The conformance transport is a recorded, network-free response from the reviewed official API URI; it proves the governed LIVE contract and deterministic admission, not that the test suite performed a live network request. FederalRegister.gov is explicitly labeled as not the official legal edition, and the corresponding govinfo.gov PDF is retained as the official-format verification reference. That frozen P2C packet creates no Signal, Shift, Brief, Decision, Outcome, feedback, delivery, publishing, persuasion, or external action.

The additive AI Command Center packet now closes the next bounded journey without mixing it into P2C. Its JSON-only five-module pack maps two exact official publication lineages to one stable AI policy entity: Federal Register document 2026-11415 for Executive Order 14409 and the White House GOLD EAGLE release of 2026-07-14. The latter reports later implementation activity rather than republishing the order. Recorded, network-free transports admit both as visibly LIVE records, then the public categorical bridge derives directive_issued → implementation_reported, one Shift, and one routed Signal. A LIVE four-member Case freezes both exact observations and the derivation; a LIVE Reality Brief cites both source snapshots across five claims. This proves two independent publication roots, not independent non-government corroboration, legal effect, implementation completeness, policy success, or current network freshness.

The accepted stable identities are pack_ir:0679116c6b5e02cd021b085ba2d80cf2, shift:d6164dce0b01e64bf8fa74bead718f77, signal:2c76e695b4b6d36af3f56fee5fdc1b20, case:5a11b2d54ce1a8402bb038415e533cbc, and brief:d7d240147e48c3aef6a4a8e7ec481121. The proof store contains 16 LIVE records and zero PREPARED records; the synthetic PREPARED packet remains byte-frozen and is not reused.

P2D adds the first complete LIVE conflict-and-correction packet. A separate inert six-module pack maps four dated publications about asteroid 2024 YR4 from two official claimant roots: NASA's 2.3% and later 0.004% Earth-impact estimates, and ESA's 1.8% and later 0.001% estimates. Recorded, network-free captures become four exact LIVE source transactions. The configured numeric bridge first routes the dated NASA/ESA estimate divergence into a historical Case and two-citation Reality Brief, then routes both same-source downward revisions into a corrected Case and four-citation Brief.

The later records do not rewrite the earlier ones. Two explicit LIVE supersession Observations name the exact same-source predecessors, and two additive impact projections enumerate five affected resources and seven affected historical Brief claims apiece while disclosing two unaffected resources. Impact means dependency, not falsehood. The original Brief reopens byte-identically after both corrections. Its status projection has two source families; the corrected projection has four raw-observation roots, but each same-lineage supersession collapses to its predecessor and adds no corroboration. The corrected corroborated claim is supported by exactly two claimant families, NASA and ESA—not by treating before/after records as independent.

The accepted P2D identities are pack_ir:bb400cc0652622b43c01504e651110e0, historical brief:c3549af0262b100ca65024ee19cbae6e, corrected brief:806d69d8e41f83f93ee3dc10f58f0d16, and corrected status projection brief_derivation_family_status_projection:2acb48dd15e625c4b729ecef484b0209. The proof store contains 44 LIVE records and zero PREPARED records. Separate official publication roots do not prove independent measurements: both agencies describe a shared international observation ecosystem.

P2E materializes the next user-owned LIVE-orientation boundary without claiming a scheduler or autonomous sensing runtime. Public ACE 0.5.0 contracts construct one exact Monitor (monitor:bbb0e5ab246c6d1f08f7669226db6873), principal/persona binding (persona_binding:1349e5c7f78270dd5c8e267b582a1399), and record-only Subscription (subscription:3bb149effa91c5fdd224780ae0a455ee) over the accepted planetary-defense activation. Six explicitly requested windows append and exactly replay initial orientation, no-change suppression, both visible corrections, owner pause/resume, and terminal revocation through the generic services merged in Core #94. WI-CR-007 and WI-CR-008 are closed for merged-source acceptance. The composed path preserves four captured NASA/ESA Observations through three Shifts, three Signals, two Cases, and the original cited Reality Briefs: 62 LIVE records and zero PREPARED records. World adds no private substitute, network request, delivery, schedule, or action. Public-index reproduction of the new Core contracts remains pending a versioned Core release.

P2C2 completes that sensing path across the Core + Intelligence + Domain boundaries. A new declarative monitor pack admits two exact FCC records — document 2026-15932 published August 6 and document 2026-16197 published August 7 — as successive LIVE snapshots of one stable monitor entity. Its configured categorical detector produces an official_publication_change Shift, routes an official_publication Signal, and invokes Core-governed reasoning to produce one LIVE Reality Brief with six claims and two exact citations. A named human Decision then authorizes only a create-only workspace export; a second exact human review precedes the effect, and separate verification and promotion receipts follow it. Replay performs neither a second reasoning call nor a second file effect.

This is a governed export proof, not autonomous publishing. The pack contains no executable code or action authority; the export is provided by Core's separately packaged reference adapter. The two source responses are exact official public records under recorded transport, so P2C2 proves the complete deterministic product journey but still does not claim network freshness at test time.

P2C3 compares two exact reviewed exports of the cited Brief with two reviewed exports of a World-owned source-only control. The frozen product measure is the fraction of the two required admitted Observation identities preserved in each exported artifact: treatment scores 1.0, 1.0; control scores 0.0, 0.0. Under the declared minimum of two matched pairs and useful-effect threshold of 0.5, Core + Intelligence classifies the exact result as useful and appends a promote proposal that is non-effective, non-selectable, and requires separate human review. Historical replay performs no reauthorization.

That measure is structural citation coverage. It is not a human-benefit measure, causal estimate, general Brief-quality score, or network-freshness proof. The proposal is not applied. World 0.10.0 binds this journey to the public ace-core>=0.6.0,<0.7 contract without moving World vocabulary, source policy, or connector code into Core.

P2C4 submits that exact proposal to a separate governed review path. A named, authenticated World reviewer records reject with no_action: the bounded useful classification and promote proposal remain immutable, and no governed head changes. The rationale rejects broader promotion because structural coverage does not establish citation correctness, general Brief quality, human benefit, causality, or live freshness. Exact replay returns the historical Decision without new authorization. This is explicit disposition, not reclassification or proposal application.

P2C5 adds a distinct product-owned outcome rather than broadening the structural score. The named principal principal:world-citation-correctness-reviewer records an exact immutable review over the Brief, its one cited claim, the exact two citation identities, and the two admitted official Observation references. Each Core Outcome names that exact result. A matched negative control retains both citation identities but swaps the two publication dates in the claim: treatment and control both have 1.0 coverage, while correctness is 1.0 and 0.0 respectively across two pairs. The domain-neutral evaluator classifies that bounded difference as useful and still emits only a non-effective proposal requiring separate review.

This candidate establishes exact review provenance and sensitivity to one semantic corruption. It does not establish reviewer infallibility, current network freshness, source independence, general Brief quality, causal impact, or human benefit. Citation review vocabulary and policy remain in World; Core and Intelligence receive only the exact immutable observed-result coordinate.

P2C6 adds an exact contradiction-attention outcome over the same recorded official sources. One candidate states the admitted document/date facts exactly and one swaps the dates. Treatment alerts only on the contradiction; an inverted-routing control alerts only on the valid comparator. Both emit exactly one alert, so raw alert volume cannot explain the 1.0 treatment/control quality difference. The independently recorded review exposes contradiction recall, false-alert rate, one valid silence, confusion counts, exact Brief/Observation provenance, policy identity, and limits.

This source-checkout candidate demonstrates one frozen challenge, not a live observed public conflict or a population false-alert rate. It does not establish autonomous discovery, network freshness, correction quality, general Brief quality, causality, or human benefit. The resulting proposal remains non-effective and unapplied.

P2C7 uses an actual explicit FCC correction pair under recorded replay. Document 2021-10670 names and corrects 2020-28779; treatment and delayed control preserve that same exact linkage, corrected instruction, original immutable Observation, and reviewed workflow. Against a frozen 600-second World target, treatment delay is 300 seconds and control delay is 21600 seconds. Exact independent reviews become Core Outcomes, and the bounded two-pair difference is useful while the proposal remains non-effective and unapplied.

The recorded availability and detection instants are test coordinates. The suite performs no network access and does not establish live monitoring, network-arrival latency, population delay performance, legal truth, calibration, general Brief quality, causality, or human benefit.

P2C8 creates actual prior, treatment, and control BriefV1Alpha1 records over that same source pair. Both revisions remove the stale instruction claim, add the same exact correction claim, retain three claims, and cite the same original/correction sources. Treatment preserves the exact identities of both unaffected claims; the paraphrase-drift control preserves neither. Independent reviews record every expected, preserved, drifted, and unexpected claim identity, producing scores of 1.0 and 0.0 across two matched pairs and another non-effective proposal.

This freezes a World product rule over one recorded pair. It is not live revision, a general semantic-equivalence engine, population stability, calibration, causality, or human benefit.

P2C9 freezes one exact probabilistic scoring boundary over that recorded correction. After the original Observation is admitted but before the correction is available, treatment and control record probabilities of 0.75 and 0.25 for the same explicit-correction event and complete the same reviewed workflow. Only then is the exact correction Observation admitted. Independent reviews derive the binary result and single-event Brier quality contributions of 0.9375 and 0.4375; the two-pair difference is useful and still emits only a non-effective proposal.

The forecast artifacts contain no correction identity or result material, and every reviewed forecast Action completes before the result becomes available. The probabilities are declared fixture inputs, not ACE outputs. This is exact forecast/result scoring, not a historically contemporaneous forecast, model-skill finding, empirical calibration curve, population reliability estimate, causal claim, or human-benefit finding.

P2C10 independently reproduces correction-quality measurement over the U.S. Bureau of Labor Statistics public errata family. The recorded July 1, 2025 JOLTS release USDL-25-1087 and BLS's July 2 correction are admitted as distinct immutable Observations. Treatment and control preserve both source identities, the exact correction link, two reviewed Actions each, and the same matched conditions; treatment renders the required minus sign in −39,000, while control retains the reported pre-correction form. Exact reviews score 1.0 and 0.0, so the unchanged domain-neutral contract classifies the bounded two-pair difference as useful and emits only a non-effective proposal.

The historical wrong form is derived from BLS's explicit statement that the sentence required a missing minus sign; the currently archived release is already corrected. This is recorded replay over one correction, not live monitoring, statistical validation of JOLTS, population correction performance, causality, or human benefit. No Domain Pack, connector, or Core contract changes.

What the public World proof demonstrates

Generate a self-contained visual Reality Brief and its exact machine-readable backing data:

PYTHONPATH=/path/to/domain-world-intelligence:/path/to/ace-core \
  /path/to/ace-core/.venv/bin/python scripts/public_demo.py

The command writes artifacts/public-demo/index.html and demo-data.json. It runs the accepted independence and correction-impact harnesses, refuses to render if the pinned release proof drifts, and uses no external web assets.

It demonstrates four things, and only these:

1. Two publishers are not two sources. Coastal Wire and Harborview are different publishers, but both derive from the Ledger report and collapse into one derivation family. Only the Ledger + Basin pairing satisfies the Brief's two-family corroboration rule.
2. Every statement carries its own epistemic status. The seven-label World vocabulary is validated by ACE against each statement's exact supports, not asserted by the pack.
3. Corrections append rather than rewrite. Admitting the later Ledger correction produces a governed impact projection over 11 downstream resources and 9 Brief claims, and discloses the 16 resources it found unaffected. Impact means dependency, not falsehood.
4. The result is reproducible. The prior Brief keeps its exact identity and replays byte-identically, and the two artifacts are byte-identical across runs.

Exact public identities reproduced by the demo:

• Case: case:412426eee708d56f6bda931ccf9e5d8b
• Brief: brief:25d8232c9bfa27050bdcb160fb75f06c

The page is explicitly a synthetic PREPARED / FROZEN scenario, not a live news product. It does not monitor the news, and it takes no action of its own. See docs/releases/world-intelligence-public-demo-v0.7.0.md for the 0.7.0 release boundary and full identity list.

Connector boundary

This repository ships an inert Domain Pack. The Federal Register connector is a separately versioned executable artifact, ace-ext-world-federal-register-source 0.2.0, with its own review boundary:

• it is not a dependency of the Domain Pack, and no extra reintroduces it;
• installing ace-domain-world-intelligence never installs it;
• it is not published by this repository's release workflow, which builds and uploads the root sdist and wheel only;
• its network client is opt-in and limited to the exact reviewed Federal Register URI; it validates the artifact identity, source type, HTTPS URI, empty redirect chain, absence of credentials, expected content type, bounded payload, globally routable address attestations, DNS-rebinding protection, and monotonic operation times;
• to exercise it, put it on the path explicitly. Omit it and the P2C modules skip rather than fail.

Rebuild the connector reproducibly:

cd adapters/federal_register_source
SOURCE_DATE_EPOCH=1735689600 uv build --wheel

A production host must supply and review a real transport that enforces address validation and rebinding protection throughout use. The shipped conformance material does not do that for you.

Verification

With $REPO, $ACE, $PYTHONPATH, and $PY exported as in Develop from a source checkout:

# Complete domain suite, including P2C2, AI Command Center, and P2D correction: 90 passed
export PYTHONPATH="$PYTHONPATH:$ACE/adapters/reference_workspace_action/src"
$PY -m pytest -q

# Connector fail-closed unit suite: 80 passed
$PY -m pytest adapters/federal_register_source/tests -q

# Individual acceptance harnesses
$PY scripts/p2a_compile_acceptance.py
$PY scripts/p2b_scenario_acceptance.py --negative
$PY scripts/p2b_prepared_replay.py
$PY scripts/p2b_case_brief.py
$PY scripts/p2b_status_case_brief.py
$PY scripts/p2b_independent_case_brief.py
$PY scripts/p2c_federal_register_live_acceptance.py
$PY -m scripts.ai_command_center_live_acceptance
$PY -m scripts.p2d_live_conflict_correction

# Complete official-record -> Shift -> Signal -> Brief -> reviewed export journey
WORKSPACE=$(mktemp -d)
$PY -m scripts.p2c2_governed_reality_brief "$WORKSPACE"

# Candidate measured Outcome -> governed-feedback extension (requires Core PR #88 source)
$PY -m scripts.p2c3_measured_feedback "$WORKSPACE"

# Stacked candidate: explicit reject/no-action review of the exact proposal
$PY -m scripts.p2c4_reviewed_impact_disposition "$WORKSPACE"

# Stacked candidates: correctness, attention, correction delay, revision stability, and forecast scoring
$PY -m scripts.p2c5_citation_correctness_outcome "$WORKSPACE"
$PY -m scripts.p2c6_contradiction_attention_outcome "$WORKSPACE"
$PY -m scripts.p2c7_correction_detection_delay_outcome "$WORKSPACE"
$PY -m scripts.p2c8_correction_revision_stability_outcome "$WORKSPACE"
$PY -m scripts.p2c9_forecast_calibration_outcome "$WORKSPACE"
$PY -m scripts.p2c10_independent_correction_reproduction "$WORKSPACE"

The 0.10.0 gates are reproducible through the locked environment, as CI does. The P2C2 action test runs when the independently packaged Core reference adapter is installed:

uv sync --frozen --no-install-project
uv run --no-sync python -m pytest                         # released + candidate-compatible gates
uv run --no-sync python -m pytest tests/test_release_contract.py  # publishable-identity gate
uv run --no-sync python -m pytest adapters/federal_register_source/tests  # connector: 80 passed

The released 0.9.0 baseline, measured with public ace-core==0.5.0 from PyPI and the independently packaged reference adapter from the Core v0.5.0 GitHub release, was 83 passed for the domain suite and 26 passed for the connector. With the AI command-center candidate integrated over that baseline, the same environment passes 87 domain and 62 connector tests. The public demo reproduces case:412426eee708d56f6bda931ccf9e5d8b and brief:25d8232c9bfa27050bdcb160fb75f06c, and its two artifacts are byte-identical across runs. The merged P2D baseline passed 90 domain tests, 80 connector tests, and all 7 release-contract tests. Its wheel contained exactly 55 JSON resources, nine from the planetary-defense pack, and no executable code. The verified P2E runtime candidate passes 102 domain, 80 connector, and all 7 release-contract tests. Its wheel contains exactly 59 JSON resources, 13 from the planetary-defense pack, including exactly four P2E conformance resources and no executable payload. A checkout-free install of the Core #94 candidate wheel reproduces the pinned Monitor, PersonaBinding, Subscription, five lifecycle transitions, six sensing receipts, and full captured LIVE lineage from site-packages without importing the Core checkout.

The exact identities, negative cases, and artifact proofs are recorded in docs/audits/world-intelligence-p2a-2026-08-06.md and docs/audits/world-intelligence-p2b-2026-08-06.md, with the interpreter and Case proof in docs/audits/world-intelligence-p2b-prepared-replay-2026-08-07.md and the governed Case-bound Brief proof in docs/audits/world-intelligence-p2b-case-brief-2026-08-07.md, the per-statement status proof in docs/audits/world-intelligence-p2b-status-case-brief-2026-08-07.md, and the independent-corroboration proof in docs/audits/world-intelligence-p2b-independent-case-brief-2026-08-07.md. The governed official-source admission proof is recorded in docs/audits/world-intelligence-p2c-federal-register-live-2026-08-07.md. The opt-in external transport observation is recorded in docs/audits/world-intelligence-p2c-network-transport-2026-08-10.md. The complete governed product-journey evidence is recorded in docs/audits/world-intelligence-p2c2-governed-reality-brief-2026-08-10.md. The two-lineage LIVE Shift → Signal → Case → cited Brief proof is recorded in docs/audits/world-intelligence-ai-command-center-live-2026-08-10.md. The source-checkout measured-feedback candidate is recorded in docs/audits/world-intelligence-p2c3-measured-feedback-2026-08-10.md. The latest independent-source correction candidate is recorded in docs/audits/world-intelligence-p2c10-independent-correction-reproduction-2026-08-10.md. The bounded installed-Core artifact convergence packet is frozen in docs/design/world-measured-intelligence-release-convergence-work-packet-v1.md. It generates one canonical machine-readable P2C10 result without importing Core from a checkout; it is candidate evidence, not a release or SI4 pass. The original point-in-time stacked convergence audit and the later live-main convergence audit bind their exact built-artifact hashes and byte-reproducible convergence-v1.json. The live-main packet also proves coexistence with the separately merged AI Command Center lineage work without changing the measured-feedback contract or silently applying its proposal. The additive P2D main reconciliation audit records source-level coexistence with the later merged LIVE correction packet without relabeling the earlier artifact hashes. The subsequent merged-Core convergence audit rebuilds all four candidate wheels from exact merged Core and World sources and freezes the updated byte-reproducible canonical result. It remains draft artifact evidence, not a release or SI4 pass. The LIVE conflict, correction, supersession, and per-claim status proof is recorded in docs/audits/world-intelligence-p2d-live-conflict-correction-2026-08-11.md. The user-owned static intent and repeated-window contract boundary is recorded in docs/audits/world-intelligence-p2e-user-owned-live-orientation-2026-08-11.md.

Release-level scope and evidence are recorded in docs/releases/world-intelligence-p2c2-v0.9.0.md, and version history in CHANGELOG.md.

Guardrails

• A publisher is not assigned a hidden universal truth score.
• Repetition and syndication are not independent corroboration.
• Claims retain attribution.
• Corrections append state rather than rewriting history.
• Inference and scenario material remain explicitly labeled.
• Persona routing never changes an evidence status.
• No political persuasion, voter targeting, or autonomous publishing.
• External effects require an exact Core-governed Decision, human review, post-effect verification, and promotion; the reference proof is limited to one create-only workspace file.

World Intelligence does not watch live news feeds, decide anything on its own, or publish or deliver content autonomously. Every LIVE capture demonstrated here is one explicitly requested, governed, read-only retrieval. P2C2's only effect is an explicitly authorized and reviewed create-only local workspace export followed by separate verification and promotion.

Next: turn one-shot governed sensing into explicit user-owned monitors and subscriptions while preserving correction visibility, suppression controls, and source-family independence. That work may not add publishing, delivery, persuasion, or other external-action authority to a Domain Pack.

Roadmap and project status

The World Intelligence roadmap owns current domain direction. Detailed packet history remains in docs/world-intelligence-roadmap-status-2026-08-06.md, and release history is in CHANGELOG.md. P2C4 demonstrates a separately authorized reject/no-action disposition of the P2C3 proposal without effective state change. P2C5 adds an independently reviewed citation-correctness Outcome and a citation-preserving semantic negative control. P2C6 adds exact contradiction recall, false-alert rate, equal alert-volume control, and valid silence under one frozen recorded-source challenge. P2C7 adds exact handling of one explicit recorded correction pair, preserves the prior record, and measures frozen-replay detection delay against a product target. P2C8 measures exact unaffected-claim identity preservation across real Brief contracts while holding correction semantics, source coverage, and claim count constant. P2C9 adds an exact withheld-result forecast record and derives a single-event Brier contribution while explicitly withholding any population-calibration or model-skill claim. P2C10 reproduces correction-quality measurement over an independently sourced BLS erratum without changing the shared contracts. The next bounded measurement work is independent Market reproduction. P2D separately demonstrates LIVE multi-source conflict, correction, supersession, and preserved historical artifacts under recorded transport. The next bounded World product work is explicit user-owned monitoring and subscription state. Separately reviewed opt-in network transport remains independent work. None of these steps may add autonomous publishing, delivery, persuasion, or action authority to a Domain Pack.

Community and security

• Contributing
• Security policy
• Code of Conduct
• Issues

License

Apache-2.0. See LICENSE and NOTICE. Existing work is copyright Edwin Amirian; contributors retain copyright in their contributions and license them under Apache-2.0.