Project Shadow 1.0.1 contains no Myth package. Generic Myth v0.2.0 and Full-Canon Myth v0.3.5 are separate optional companions; both default off, neither is required by R1, and neither can authorize action or change an R1 result. Civic QA remains self-service and unofficial; publication transfers no authority and is not operational deployment, efficacy, safety, certification, a legal-compliance finding, permission to act, or a global green.
Six real questions. Six honest evidence boundaries.
These demonstrations show how anyone can run Civic QA across Congress, candidates, local services, healthcare, public-benefits AI, and federal agency performance. They are current-source teaching inspections—not official findings.
EVERY CASE BELOWWORKED DEMONSTRATION · UNOFFICIAL · PUBLIC SURFACE 2026-08-14 · EVIDENCE CUTOFFS ITEM-SPECIFIC · PHILLIP REVIEW PENDING
Documented is not validated. Missing official public data remains a finding about the evidence boundary—not permission to invent an outcome.
SIX LAUNCH CASES
Change the institution. Keep the inspection discipline.
Each case names the exact question, what official evidence currently supports, what it does not establish, the deterministic gate, the next evidence action, and direct sources.
01 / CONGRESS / ETHICS / LEGISLATION
Has Congress enacted a real stock-trading ban?
Did Congress replace disclosure-based controls with an enforceable ban covering members, spouses, dependent children, ownership, trading, anti-evasion, independent enforcement, and meaningful penalties?
KNOWN / DOCUMENTED
The STOCK Act prohibits use of nonpublic official information and requires disclosures; it is not a blanket ownership ban.
House and Senate ethics processes rely heavily on financial and periodic transaction disclosures.
S.1498 and H.R.1908 were pending—not enacted—when checked.
UNKNOWN / NOT ESTABLISHED
No single cross-chamber machine-readable audit was located for filing completeness, timeliness, late fees, referrals, and enforcement outcomes.
NEXT EVIDENCE ACTION
Hold claims of a comprehensive ban. Stop publication of any statement that the pending proposals are enacted.
02 / INDIANA 6TH / CANDIDATE ACCOUNTABILITY
Can Shreve and Wirth be compared through the same questions?
Do both candidates provide usable, source-linked, measurable commitments across access, pocketbook, power, and delivery—and does incumbent delivery evidence map to those commitments?
KNOWN / DOCUMENTED
Indiana’s reporting portal shows Wirth winning the Democratic primary but retains an unofficial-results caveat.
The House Clerk identifies Shreve as the incumbent and records his current Appropriations assignment.
No official same-question candidate questionnaire or complete claim-to-record crosswalk was located.
Evidence asymmetry is not evidence of deception; a challenger does not have an incumbent congressional record.
NEXT EVIDENCE ACTION
Ask identical questions, label unanswered fields identically, and keep campaign statements separate from official records.
03 / INDIANAPOLIS / REQUESTINDY / SERVICE OUTCOME
Does “Closed” mean the resident’s problem was fixed?
For one service class and time window, does an administratively closed RequestIndy record prove corrective action, resident-confirmed resolution, and durable non-recurrence?
KNOWN / DOCUMENTED
Indianapolis publishes a live service-request FeatureServer with request, service, department, location, requested, updated, closed, status, and origin fields.
Status and timestamps can support workflow analysis.
UNKNOWN / NOT ESTABLISHED
The inspected public schema does not expose standardized disposition, corrective-action proof, verification method, resident confirmation, reopen reason, or SLA.
Public demonstrations should redact residential addresses.
NEXT EVIDENCE ACTION
Hold outcome and effectiveness claims until a defined sample has closure reasons, field or documentary verification, and recurrence checks.
04 / INDIANA MEDICAID / PRIOR AUTHORIZATION
Do documented deadlines prove timely access to care?
Do 2026 non-drug prior-authorization decisions meet required timelines, provide a usable appeal, and produce measurable access rather than merely a processed decision?
KNOWN / DOCUMENTED
Indiana documents a 72-hour deadline for urgent requests and seven calendar days for standard requests beginning January 1, 2026.
Indiana documents free member appeals and separate health-plan and FSSA routes.
UNKNOWN / NOT ESTABLISHED
A full-year 2026 result cannot yet exist.
No centralized normalized MCO comparison was located for deadline breaches, appeal reversal, abandonment, time-to-care, and disparities.
NEXT EVIDENCE ACTION
Hold any validated-access claim. Build a denominator-consistent payer and service comparison when the data window exists.
05 / FEDERAL BENEFITS / AI PROCUREMENT
What must exist before a high-impact AI system earns use?
Should a benefits agency award or deploy eligibility, fraud, or triage AI before representative testing, impact assessment, human remedy, monitoring, rollback, and exit rights are inspectable?
KNOWN / DOCUMENTED
OMB M-25-21 treats AI materially affecting access to critical government resources or services as high-impact and requires minimum risk practices.
OMB M-25-22 addresses performance, testing, monitoring, competition, lock-in, data rights, and acquisition terms.
M-26-10 adds technology-agreement reporting and machine-readable internal oversight requirements.
UNKNOWN / NOT ESTABLISHED
Award pages and AI inventories often omit full contracts, QASPs, evaluation datasets/results, impact assessments, waivers, subgroup outcomes, and appeal evidence.
NEXT EVIDENCE ACTION
Hold at procurement. Stop deployment or automated adverse action when the high-impact package, qualified human remedy, logs, or rollback proof is absent.
06 / SSA / DISABILITY / PERFORMANCE CLAIM
Improved is not the same as resolved.
Is the disability backlog improving without hiding long-tail waits, abandonment, geographic variation, accuracy, remand, reversal, and filing-to-payment time?
KNOWN / DOCUMENTED
SSA reported lower initial backlog and faster initial and hearing decisions in 2026.
SSA’s hearing goal is 270 days, and official monthly datasets explain important denominator and exclusion limits.
UNKNOWN / NOT ESTABLISHED
The public presentation does not provide one complete view of median and P90 waits, age buckets, abandonment, harm, accuracy, and time to payment.
Monthly averages for closed cases are not a complete picture of people still waiting.
NEXT EVIDENCE ACTION
Publish the headline “Improved, not resolved” and keep tail-wait, accuracy, abandonment, and variation as corrective-action items.
REQUIRED OUTPUTThese cases teach the method. None receives an official Civic QA mark until Phillip reviews the exact edition, source trail, rationale, hard stops, burden, disclosures, and correction path.
PUBLICATION GOVERNANCE
Anyone can inspect. A mark belongs only to one reviewed edition.
Official does not mean infallible or certified safe. It means the report met Civic QA's publication controls and remains challengeable with evidence.
01
SELF-DRAFT
Anyone may create, import, inspect, and export. No official mark.
02
WORKED DEMONSTRATION
A teaching inspection built from attributable evidence. Still unofficial.
03
UNDER REVIEW
Scope, sources, rationales, hard stops, burden, disclosure, and correction path are challenged.
04
REVIEWED
Phillip approves one exact edition, scope, date, and receipt for the official mark and listing.
05
REOPENED / CORRECTED
Material new evidence suspends the mark until a corrected edition passes review again.
06
WITHDRAWN
A public tombstone preserves the reason, date, and prior receipt. Nothing vanishes silently.
USE THE SAME WORKBENCH
Load any demonstration, inspect every field, change the evidence, and export your own draft.