HXR: Current Project Status

Current stage: founder-led Architectural Build Publication Mode within Private Alpha. HXR accepts a limited number of real-world submissions and uses them to develop and inspect the canonical records, evidence graph, Evaluations, and real public frontend. HXR is still a one-person project, not a staffed agency or mature certification system. Human Experience Reform is now the active public identity. The August 18, 2026 UXR-to-HXR cutover is complete. The current MAIN Shopify theme includes the verified HXR Home, Documentation, Cases, Incident, Evaluation, support, reviewer, development, and navigation presentation. Legacy `UXR-...` identifiers, `/blogs/uxr` routes, and internal `uxr_*` keys remain intentionally stable. Intake is not automatic publication. A submission does not automatically become an Incident, Evaluation finding, vote, prevalence claim, or public record. When Sean explicitly authorizes publication of a bounded Case or Evaluation, the ordinary public projection may be completed with privacy, permitted-use, provenance, canonical-linkage, and write/readback safeguards.

Current Operational Surface

• Canonical Notion records and a controlled Shopify public projection are operational.
• HXR Documentation, HXR Cases, the projection-native Case Library, public Incident and Evaluation templates, filters, search, relationships, and public routes are operational.
• HXR has reached a public-facing outside-review milestone. Since the August 18 UXR-to-HXR rename and cutover, the project now presents one coherent public HXR identity backed by real published Incidents and Evaluations, a live Case Library, reconciled documentation projections, continuity-preserving routes, and current journalist/reviewer entry points. This is readiness for inspection and discussion, not a claim that HXR has become a staffed agency, validated scoring authority, or mature scientific institution.
• The three earliest canonical public Evaluations, `UXR-EVAL-0001`, `UXR-EVAL-0002`, and `UXR-EVAL-0004`, received an August 25 compatibility audit against current HXR field semantics and Evidence Graph rules. Their public/search framing, positive/negative or explicit non-findings, evidence ceilings, unresolved questions, agency dimensions, reform posture, privacy/controls where applicable, and contribution states were brought forward without renumbering legacy IDs or manufacturing new evidence. Their existing public projections were updated and independently read back within the current 40-field Evaluation projection rather than expanding that schema.
• The post-publish HXR test confirmed the live unified theme and repaired generic-template assignments on 10 documentation articles and five public-record articles.
• Core current-identity documents, including the Copyright Notice, Constitutional Integrity Pledge, Brand and Usage Guidelines, Research Notebook, and Development Registry, are synchronized with append-only revision records.
• The documentation projection identity map was re-audited against Shopify by stable document key. Forty-five incorrect Notion `Shopify Projection ID` values were corrected; the mapped corpus now contains 73 unique Notion→Shopify documentation identities with no duplicate projection IDs. One newer document, How Your Contribution Can Matter, legitimately remains without a v2 Projection ID and must be created through the governed publication workflow rather than assigned a guessed ID.
• Seventeen already-live documentation projections received bounded HXR title/short-summary corrections with independent readback. That operation deliberately did not pretend their bodies, source hashes, revisions, routes, or sync timestamps had completed a full canonical delta.
• The automated hourly Documentation Publisher is currently disabled and held in Shadow / Dry Run after a September 2 rich-text transform regression. `UXR-DEV-34` defines the fail-closed re-enable tests. Until those tests pass, governed manual projection with deterministic hashing, exact admin readback, live HTML inspection, legacy Article parity where applicable, and append-only sync logging is the active documentation synchronization path.
• Private evidence, contributor-sensitive material, workbench records, internal notes, and non-public governance data remain outside the public projection.

Minimal Evaluation Contribution Pilot

The contribution pilot is deliberately small. The canonical and public Evaluation projections now contain:

• Contribution Intake State;
• Public Contribution Prompt; and
• Minimum Matching Context.

The first two configured pilots are the DMHC timely-access remediation Evaluation and the California health-plan specialist-access Evaluation. Five Amazon bulk-listing Evaluations, `HXR-EVAL-0042` through `HXR-EVAL-0046`, now run an additional bounded manual-email pilot on the live theme. Each exact-symptom Evaluation keeps one Add Your Experience doorway, offers a 10-second prefilled confirmation, and links back to the single Incident hub. The five Evaluations remain separate because their remedies and acceptance tests are independent. Current intake is human-operated. The secure contribution flow and automated public-to-ledger bridge are not live. A person sends a story or bounded confirmation to Sean/HXR through the available human contact route. Sean or another expressly authorized operator then opens the current intake agent and manually creates or reconciles the private Intake Session, privacy/use state, candidate matches, evidence relationships, and plain-language summary. Intake Sessions now have fields for acknowledgment state, sent time, a private outbound-delivery reference, and the stable receipt reference communicated to the contributor. Those fields make receipt parity structurally available; no real contributor transaction had yet completed acknowledgment plus full canonical readback as of September 2, 2026. The existing Amazon email actions remain non-secure first contact only and should not carry identifiers, confidential material, or sensitive evidence. This manual process does not promise publication, automatic notifications, an organization response, or a particular outcome. Organization responses use an adopted manual bridge with a separate ledger. A representative's message is transport until an authorized operator preserves the original received time and payload, creates or reconciles the private `HXR-RESP-N` record, applies private defaults, records authority and granular permissions, assigns append-only claim labels and dated dispositions, sends a separate stable-reference acknowledgment, and reads back the transaction. The necessary fields and procedure are structurally available, but a real organization-response transaction had not yet completed that acknowledgment and full canonical readback as of September 2, 2026. The public form is not a secure evidence portal or automatic receipt. All service times are capacity-qualified, good-faith, noncontractual targets. For target calculation, a business day is Monday through Friday excluding U.S. federal holidays, measured in Pacific Time; a message received after 5:00 p.m. Pacific Time is treated as received on the next business day.

Healthcare Pilot Canonical State

The healthcare domain has completed a full canonical reconciliation against the current Workbench and Evidence Graph. The July 29–August 6 cardiology sequence is now represented as one active bounded Incident rather than several mechanism-level child Incidents. Three former cardiology mechanism slices are retired for historical continuity, while one distinct Covered California enrollment/remedy occurrence remains separate because its objective is different. Seven healthcare Evaluations remain in Active Evidence Collection and receive evidence-specific contribution edges from those two bounded occurrences. This lets the same cardiology occurrence partially support, narrow, contextualize, or support different functional Evaluations without being counted as several independent incidents. Only the California health-plan timely-access Evaluation and the DMHC remediation Evaluation are configured as the two Phase 1 contribution pilots; the other five remain Not Configured. The healthcare audit also formalized seven Evidence Items, controlled-function attribution, explicit non-findings, unresolved questions, and acceptance tests. Public healthcare projections were synchronized to the canonical model, including removal from public presentation of the retired cardiology mechanism slices. This domain pass did not change the shared Add Your Experience, intake, attachment, matching, portal, notification, or Impact Receipt architecture.

Current Technical Constraint

Shopify's public Evaluation definition is at its 40-field maximum. The three approved contribution fields fit. `Material Unresolved Questions` does not fit as an additional field without deleting or redesigning an existing projection field. For the two healthcare pilot records only, the theme projects the exact canonical unresolved-question text through a clearly marked mapping. The five Amazon manual-email pilots use only the three approved contribution fields and do not extend that fallback. The unresolved-question mapping must not be generalized without a separate reconciliation decision.

Immediate Next Work

1. Operate the single Add Your Experience doorway through the governed manual handoff: receive the story, open the intake agent, preserve the actual received time and privacy/use choices, and create or reconcile the private Intake Session.
2. Return a manual plain-language summary of what HXR recorded, the candidate Evaluation match, privacy/use state, what the contribution may add, what it does not establish, and the next safe step.
3. Use the adopted manual organization-response SOP for real representative contacts, including stable `HXR-RESP-N` acknowledgment, authority scope, granular permissions, append-only claim review, reconsideration, conflict controls, and readback.
4. Observe the healthcare and Amazon pilots for comprehension, effort, matchability, correction load, privacy understanding, response time, failure recovery, and preservation of positive or materially different experiences. Build automation only if observed use justifies it. A safer evidence-transfer method remains required before asking for sensitive records.
5. Continue only bounded visual/current-language cleanup where a confirmed defect remains. The structural desktop/mobile review of HXR Home, Documentation, Cases, one Incident, and the DMHC Evaluation has passed; remaining work should not become another global rename or route churn.
6. Keep the automated Documentation Publisher disabled and in Shadow / Dry Run until `UXR-DEV-34` passes. Process any authorized canonical delta through governed manual projection with deterministic hashing, exact admin readback, live HTML inspection, legacy Article parity where applicable, and append-only sync logging. Do not infer full body synchronization from the earlier 17 title/summary-only corrections.

Still Under Validation

The first 12-fixture manual regression baseline produced eight passes and four failures. It was single-grader, self-graded, sequential, and did not place an executable publication broker in the decision path. HXR froze v1 before remediation, added candidate/source-resolution records, adopted a no-manufactured-attribution invariant, made linked result rows authoritative, added idempotent write identity, and drafted successor coverage. Broker enforcement, automated scoring, mature formal release control, broad public intake, automated matching, subscriptions, notifications, portals, and a persistent Contribution Impact Receipt remain development work. They are not prerequisites for ordinary founder-authorized frontend iterations, and none should be built merely because it sounds useful.