Issue explicit stop orders
Name every active system and every revoked target identifier. Record when the order was acknowledged and preserve a hash of the stop evidence.
R45 · Remedy execution · five fictional demonstrations
Make the old judgment stop acting.
A correction is not real until the old result stops acting. A revoked label must lose operative force everywhere it traveled. Auditability must not become permanent stigma.
Every person, institution, system, allegation, order, and event on this page is fictional. The records are constitutional and technical design demonstrations. They are not evidence that a real institution uses these practices, and they are not legal advice.
Direct answer
A remedy is an executed state change, not a corrected sentence.
A notice saying “we corrected the record” is inadequate while an old score still blocks a payment, feeds a watch queue, survives in a cache, trains a derived feature, or travels in an export. R45 therefore treats correction as a verifiable sequence: authorize, stop, recompute, propagate, prove, and notify.
Rebuild from permitted inputs
Exclude disputed or purpose-incompatible inputs, pin the algorithm version, and publish canonical input and output hashes.
Revoke every downstream copy
Deliver explicit messages through an acyclic dependency graph. An acknowledgment must bind to the exact message and target identifiers.
Prove the stigma cannot act
Test active queries, caches, features, and training registries. A sealed audit residue remains segregated, narrow, logged, and expiry-bound.
Exactly five scenarios
Different failures, one execution discipline.
The scenario cards are summaries, not rankings. Each exact receipt is validated against the single antichrist.cx.remedy-receipt.v1 authority.
Housing hold corrected after cross-context signal separation
Purpose review established that an imported fraud signal could not lawfully or logically stand in for current housing evidence.
verified
Student support record prevented from becoming a punitive threat identity
Human review confirmed that a support referral was purpose-limited and could not become a punitive or permanent threat label.
verified
Employment-benefit result recomputed after speculative health inference removal
The remedy owner removed a speculative health inference that was unsupported by purpose-matched evidence.
verified
Public-event restriction revoked across credential and downstream watchlist-like systems
Association evidence was disaggregated from personal intent and the access restriction was formally rescinded.
partial exception
Predictive public-safety label withdrawn after base-rate and provenance review
A separate remedy review found that a predictive label overread a low-base-rate association signal and lacked source-provenance fitness for individual action.
verified
One explicit state machine
No silent skips. No silent reactivation.
The normal path is proposed → authorized → stop ordered → recomputing → propagating → verified. Rejected, expired, rolled back, and partial-exception branches remain explicit. A terminal revocation cannot quietly return to active use.
- proposed
A bounded correction request identifies the superseded receipt and target scope.
- authorized
A named human owner accepts authority and responsibility for the correction.
- stop ordered
Every current use receives an explicit stop instruction before recomputation continues.
- recomputing
The decision is rebuilt from permitted inputs under a pinned deterministic version.
- propagating
Revocation messages move through declared downstream dependencies and collect acknowledgments.
- verified
All active uses are stopped, required acknowledgments exist, subject proof is delivered, and forgetting tests pass.
Partial exception
A delivery failure remains visible, human-owned, deadline-bound, and incapable of being mislabeled as full verification.
Typed and reviewable
An emergency skip must state authority, human owner, reason, exact scope, and review deadline. Missing any field fails closed.
Canonical idempotency
The same key plus the same canonical payload yields the same result. A changed payload under the same key is rejected.
Restorable server state
Choose a receipt and a human lens.
The URL carries the selected receipt and role. The role changes reading emphasis only; it never hides or rewrites the underlying record.
What stopped acting, what changed, and when do I receive proof?
This lens changes reading order only; it hides no remedy field.
Priority fields: subject_notice · stop_orders · recomputation outputs · contest route
Antichrist.cx R45 human-readable remedy receipt
RMR-R45-005 · exact fictional receipt
Predictive public-safety label withdrawn after base-rate and provenance review
Fictional remedy demonstration only. This record describes no real person, institution, allegation, system, or event.
- Affected person
- Fictional reviewed person (SUB-005)
- Correction authority
- Fictional predictive-safety review order R45-PSA-05
- Correction trigger
- A separate remedy review found that a predictive label overread a low-base-rate association signal and lacked source-provenance fitness for individual action.
- Superseded receipt
- CTR-R44-004 · DEC-004
- Current result
- Predictive label withdrawn; current systems return no active risk identity while a narrow sealed audit proof remains.
- Human sign-off
- All declared active uses stopped and all downstream holders acknowledged revocation.
Stable relationship to R44
The remedy references the old decision; it does not copy and mutate it.
The R45 record binds to the exact R44 contestability receipt by stable identifier and canonical hash. That preserves historical accountability while keeping the corrected current state separate.
CTR-R44-004
Decision ID: DEC-004
Scope stopped: predictive public-safety label, watch queue, and derived proximity feature
R44 hash: sha256:8ab41ae675196e963fbdb03762b4e758c1fba0888aed75720a1b5cb0f8e071ca
Association signal disaggregated and restriction rescinded
Prior status: rescinded
Prior outcome: The access restriction is rescinded and all downstream copies require revocation.
Path nodes: 6
Previously declared downstream holders: 3
Plain-language execution timeline
Who changed what, under which authority, and when.
-
proposed → authorized
2026-08-30T12:10:00Z
Advance remedy after proposed evidence completed.
Accountable actor: Fictional human remedy owner · Scope: predictive public-safety label, watch queue, and derived proximity feature
-
authorized → stop_ordered
2026-08-30T12:20:00Z
Advance remedy after authorized evidence completed.
Accountable actor: Fictional operations lead · Scope: predictive public-safety label, watch queue, and derived proximity feature
-
stop_ordered → recomputing
2026-08-30T12:30:00Z
Advance remedy after stop_ordered evidence completed.
Accountable actor: Fictional operations lead · Scope: predictive public-safety label, watch queue, and derived proximity feature
-
recomputing → propagating
2026-08-30T12:40:00Z
Advance remedy after recomputing evidence completed.
Accountable actor: Fictional downstream coordinator · Scope: predictive public-safety label, watch queue, and derived proximity feature
-
propagating → verified
2026-08-30T12:50:00Z
Advance remedy after propagating evidence completed.
Accountable actor: Fictional human remedy owner · Scope: predictive public-safety label, watch queue, and derived proximity feature
Active-use stop orders
The former result loses force before the replacement is trusted.
Public-safety review queue
stopped
Ordered: 2026-08-30T12:23:00Z
Acknowledged: 2026-08-30T12:26:00Z
- LABEL-PREDICTIVE-SAFETY
- FEATURE-PROXIMITY-RISK
- QUEUE-HIGH-RISK
Stop evidence: sha256:39ed366ea1ab6bb5d77fc7b4d560636d0a08bf4938835848bb502ce3a569fd2a
Mobility screening cache
stopped
Ordered: 2026-08-30T12:24:00Z
Acknowledged: 2026-08-30T12:28:00Z
- LABEL-PREDICTIVE-SAFETY
- FEATURE-PROXIMITY-RISK
- QUEUE-HIGH-RISK
Stop evidence: sha256:19f7accc9ece864c1059f3296d0b407ca449cde3d2f8997a5714ddb7f8a96f70
Model-training exclusion registry
stopped
Ordered: 2026-08-30T12:25:00Z
Acknowledged: 2026-08-30T12:30:00Z
- LABEL-PREDICTIVE-SAFETY
- FEATURE-PROXIMITY-RISK
- QUEUE-HIGH-RISK
Stop evidence: sha256:1dea2ad396e526bf67f2ccfd129396f45e259d2aeed2fedfaebc82682912bb58
Deterministic recomputation
Remove the invalid influence, then rebuild from declared inputs.
Engine: purpose-limited deterministic rule evaluation · Version: r45-remedy-engine-1.0.0 · 2026-08-30T12:31:00Z to 2026-08-30T12:36:00Z
Purpose-matched submitted record
direct evidence retained
sha256:5e47b74c0fe6f7957e276fbcdeccd2a49b1d021ef7e6fac8247c2c6a5251c64c
Disputed or cross-context signal
LABEL-PREDICTIVE-SAFETY
sha256:021ede297fd5541a87e821b58d12d3487eefb568067ad3cb532f5e4b97c2c94e
Human correction statement
context and correction authority
sha256:0cf015d8c4b6985554969e7db10c2b4336d9fb43d2c23c091a614f666fa698e8
Version-pinned fictional policy
Fictional predictive-safety review order R45-PSA-05
sha256:51e9055123ae4ef755b1e2099d5151f1009c9f8244c3ecf92efda90bf9beab73
Predictive label withdrawn; current systems return no active risk identity while a narrow sealed audit proof remains.
sha256:42d914de286ff08bdd21a1d77c5daeba34f5907f2644e15263884cd3b55a32ee
Former operative state removed: predictive public-safety label, watch queue, and derived proximity feature.
sha256:cc02ce50522718a388752ea4fea7d0740b93eaccef0f5bc16f11045da16b9be5
Canonical input set: sha256:7bd56a636b7d4590f566ecb10e0d4208083b2c83943fccc4ba6b0ac28add4d06
Canonical output set: sha256:2024a59fa8d47cbe7e0a690937df9e9fc5496f9a76f6a0cf1f2d479662ae4ef9
Downstream revocation graph
Every holder must acknowledge the exact revoked targets.
Public-safety review queue
acknowledged
Sent: 2026-08-30T12:43:00Z · Retries: 0 · Depends on: none
- LABEL-PREDICTIVE-SAFETY
- FEATURE-PROXIMITY-RISK
- QUEUE-HIGH-RISK
Delivery receipt: DLV-05-01
Acknowledged: 2026-08-30T12:47:00Z — Revoked targets are inactive and excluded from current decision paths.
Delivery attempts
- Attempt 1 at 2026-08-30T12:45:00Z: delivered — Revocation message delivered.
Mobility screening cache
acknowledged
Sent: 2026-08-30T12:46:00Z · Retries: 0 · Depends on: REVOKE-05-01
- LABEL-PREDICTIVE-SAFETY
- FEATURE-PROXIMITY-RISK
- QUEUE-HIGH-RISK
Delivery receipt: DLV-05-02
Acknowledged: 2026-08-30T12:51:00Z — Revoked targets are inactive and excluded from current decision paths.
Delivery attempts
- Attempt 1 at 2026-08-30T12:49:00Z: delivered — Revocation message delivered.
Model-training exclusion registry
acknowledged
Sent: 2026-08-30T12:49:00Z · Retries: 0 · Depends on: REVOKE-05-02
- LABEL-PREDICTIVE-SAFETY
- FEATURE-PROXIMITY-RISK
- QUEUE-HIGH-RISK
Delivery receipt: DLV-05-03
Acknowledged: 2026-08-30T12:55:00Z — Revoked targets are inactive and excluded from current decision paths.
Delivery attempts
- Attempt 1 at 2026-08-30T12:53:00Z: delivered — Revocation message delivered.
Proof to the affected person
The subject receives a readable result and an open contest route.
Proof delivered
Predictive label withdrawn; current systems return no active risk identity while a narrow sealed audit proof remains.
Channel: Accessible fictional written and screen-reader-ready notice
Delivered: 2026-08-30T13:02:00Z
NOTICE-PROOF-05
Receipt RMR-R45-005 was delivered for SUB-005 at 2026-08-30T13:02:00Z.
sha256:aa0f7bccd9ebd97998f04670e9a0eec0fa881ce5a956040f2708b792114d8f12
Active state versus sealed accountability
Prove correction without preserving a reusable stigma.
Current authorized state only
The active zone may expose the recomputed current outcome and proof that the former targets are revoked. It may not query the sealed history to recreate a risk score.
- Current output: Predictive label withdrawn; current systems return no active risk identity while a narrow sealed audit proof remains.
- Revoked outputs remain marked revoked.
- Forgetting probes return zero active results.
SEALED-R45-005
Prove that the correction occurred and test recurrence without enabling eligibility, suspicion, or risk scoring.
Active use prohibited: yes
Queryable by active systems: no
Expires: 2027-08-30T12:00:00Z
Retained: remedy receipt id · superseded receipt id and hash · transition timestamps · revocation acknowledgment proofs · whole-receipt hash
Excluded: raw intimate content · free-text allegations · reusable risk features · unbounded identity profile
ACCESS-05-01
ACT-R45-AUDITOR-05 at 2026-08-30T13:07:00Z
Verify correction completeness and sealed/active separation.
Institutional-forgetting evidence
The old label cannot be found by active systems.
Deletion language is not enough. Each record includes a typed probe, exact target identifiers, a test result, and a canonical proof hash.
not queryable
Query active interfaces and derived-feature registries for the revoked target identifiers.
Checked: 2026-08-30T13:11:00Z · Active results: 0 · Sealed results: 1
- LABEL-PREDICTIVE-SAFETY
- FEATURE-PROXIMITY-RISK
- QUEUE-HIGH-RISK
sha256:01589bd9ab756e705d75f4449368eaf67875a5cc704b1d061f9305086dfa5242
feature deactivated
Query active interfaces and derived-feature registries for the revoked target identifiers.
Checked: 2026-08-30T13:12:00Z · Active results: 0 · Sealed results: 0
- LABEL-PREDICTIVE-SAFETY
- FEATURE-PROXIMITY-RISK
- QUEUE-HIGH-RISK
sha256:156140ae2aa99b3c1cfc8d9d218a2ae169f0581a002213fda8f7a5367a3bd12e
cache absent
Query active interfaces and derived-feature registries for the revoked target identifiers.
Checked: 2026-08-30T13:13:00Z · Active results: 0 · Sealed results: 0
- LABEL-PREDICTIVE-SAFETY
- FEATURE-PROXIMITY-RISK
- QUEUE-HIGH-RISK
sha256:820444ff58e9e2ee176458e16a85779d919fc5c43a4fe7232e69f36a2e8a6fc0
No false completion
Exceptions remain visible and human-owned.
All declared active uses stopped and all downstream holders acknowledged revocation.
Scope: predictive public-safety label, watch queue, and derived proximity feature
All active uses stopped: yes
All downstream acknowledged: yes
Signed: 2026-08-30T13:20:00Z
Replay safety
A repeated request cannot expand or mutate the remedy.
R45-IDEMP-005
The replay result is same_result at 2026-08-30T13:25:00Z.
sha256:cdb760ce7c178df80beec22c071005fe320a2f389ae249414a5897e891d104fe
Canonical replay payload: sha256:cdb760ce7c178df80beec22c071005fe320a2f389ae249414a5897e891d104fe
sha256:3038178d16a5ababd9e78a57c9d3f8843b18443f81ae076dab29b3a3b504892a
Replay result: sha256:3038178d16a5ababd9e78a57c9d3f8843b18443f81ae076dab29b3a3b504892a
Maximum-three comparison · no ranking
Compare remedy mechanics, not human worth.
Choose up to three fictional receipts. The server enforces the same maximum as the JavaScript enhancement. Selection produces no score, preference, or rank.
Predictive public-safety label withdrawn after base-rate and provenance review
State: verified
Stop orders: 3
Downstream acknowledgments: 3 / 3
Forgetting proofs: 3
Open exceptions: 0
Exact machine-readable record
Do not trust a screenshot when canonical bytes can be checked.
The PHP authority rejects malformed UTF-8, duplicate decoded keys, unsafe Unicode controls, unknown properties, broken references, invalid transitions, changed replay payloads, forged acknowledgments, graph cycles, stale versions, ambiguous nulls, and forged hashes.
Whole receipt hash: sha256:53bb7728ce9b537f92f36b5e0f0b3a83eb680485cabb90105fa0c0a390bf32a8
{
"actors": [
{
"actor_id": "ACT-R45-OWNER-05",
"contact_route": "Accessible fictional remedy desk",
"human": true,
"name": "Fictional human remedy owner",
"role": "human_remedy_owner"
},
{
"actor_id": "ACT-R45-OPERATOR-05",
"contact_route": "Fictional operations channel",
"human": true,
"name": "Fictional operations lead",
"role": "human_system_operator"
},
{
"actor_id": "ACT-R45-DOWNSTREAM-05",
"contact_route": "Fictional downstream coordination channel",
"human": true,
"name": "Fictional downstream coordinator",
"role": "human_downstream_owner"
},
{
"actor_id": "ACT-R45-AUDITOR-05",
"contact_route": "Fictional bounded audit channel",
"human": true,
"name": "Fictional independent auditor",
"role": "human_auditor"
},
{
"actor_id": "ACT-R45-ADVOCATE-05",
"contact_route": "Fictional advocacy channel",
"human": true,
"name": "Fictional affected-person advocate",
"role": "human_advocate"
}
],
"correction_trigger": {
"authorized_by_actor_id": "ACT-R45-OWNER-05",
"discovered_at": "2026-08-30T12:02:00Z",
"source_of_authority": "Fictional predictive-safety review order R45-PSA-05",
"summary": "A separate remedy review found that a predictive label overread a low-base-rate association signal and lacked source-provenance fitness for individual action.",
"trigger_id": "TRIGGER-R45-005"
},
"downstream_revocations": [
{
"acknowledgment": {
"acknowledgment_id": "ACK-05-01",
"at": "2026-08-30T12:47:00Z",
"message_id": "REVOKE-05-01",
"statement": "Revoked targets are inactive and excluded from current decision paths."
},
"attempts": [
{
"at": "2026-08-30T12:45:00Z",
"attempt": 1,
"detail": "Revocation message delivered.",
"result": "delivered"
}
],
"delivered_at": "2026-08-30T12:45:00Z",
"delivery_receipt": "DLV-05-01",
"depends_on": [],
"label": "Public-safety review queue",
"message_id": "REVOKE-05-01",
"retries": 0,
"sent_at": "2026-08-30T12:43:00Z",
"status": "acknowledged",
"subject_id": "SUB-005",
"system_id": "PSA-QUEUE",
"target_ids": [
"LABEL-PREDICTIVE-SAFETY",
"FEATURE-PROXIMITY-RISK",
"QUEUE-HIGH-RISK"
]
},
{
"acknowledgment": {
"acknowledgment_id": "ACK-05-02",
"at": "2026-08-30T12:51:00Z",
"message_id": "REVOKE-05-02",
"statement": "Revoked targets are inactive and excluded from current decision paths."
},
"attempts": [
{
"at": "2026-08-30T12:49:00Z",
"attempt": 1,
"detail": "Revocation message delivered.",
"result": "delivered"
}
],
"delivered_at": "2026-08-30T12:49:00Z",
"delivery_receipt": "DLV-05-02",
"depends_on": [
"REVOKE-05-01"
],
"label": "Mobility screening cache",
"message_id": "REVOKE-05-02",
"retries": 0,
"sent_at": "2026-08-30T12:46:00Z",
"status": "acknowledged",
"subject_id": "SUB-005",
"system_id": "PSA-MOBILITY",
"target_ids": [
"LABEL-PREDICTIVE-SAFETY",
"FEATURE-PROXIMITY-RISK",
"QUEUE-HIGH-RISK"
]
},
{
"acknowledgment": {
"acknowledgment_id": "ACK-05-03",
"at": "2026-08-30T12:55:00Z",
"message_id": "REVOKE-05-03",
"statement": "Revoked targets are inactive and excluded from current decision paths."
},
"attempts": [
{
"at": "2026-08-30T12:53:00Z",
"attempt": 1,
"detail": "Revocation message delivered.",
"result": "delivered"
}
],
"delivered_at": "2026-08-30T12:53:00Z",
"delivery_receipt": "DLV-05-03",
"depends_on": [
"REVOKE-05-02"
],
"label": "Model-training exclusion registry",
"message_id": "REVOKE-05-03",
"retries": 0,
"sent_at": "2026-08-30T12:49:00Z",
"status": "acknowledged",
"subject_id": "SUB-005",
"system_id": "PSA-TRAIN",
"target_ids": [
"LABEL-PREDICTIVE-SAFETY",
"FEATURE-PROXIMITY-RISK",
"QUEUE-HIGH-RISK"
]
}
],
"fiction_notice": "Fictional remedy demonstration only. This record describes no real person, institution, allegation, system, or event.",
"fictional": true,
"final_signoff": {
"actor_id": "ACT-R45-OWNER-05",
"all_active_uses_stopped": true,
"all_downstream_acknowledged": true,
"at": "2026-08-30T13:20:00Z",
"conclusion": "All declared active uses stopped and all downstream holders acknowledged revocation.",
"scope": "predictive public-safety label, watch queue, and derived proximity feature"
},
"forgetting_evidence": [
{
"checked_at": "2026-08-30T13:11:00Z",
"evidence_id": "FORGET-05-01",
"proof": {
"active_result_count": 0,
"probe_id": "FORGET-PROBE-05-01",
"query": "LABEL-PREDICTIVE-SAFETY",
"scope": "active system interface only",
"sealed_result_count": 1
},
"proof_hash": "sha256:01589bd9ab756e705d75f4449368eaf67875a5cc704b1d061f9305086dfa5242",
"result": "not_queryable",
"system_id": "PSA-QUEUE",
"target_ids": [
"LABEL-PREDICTIVE-SAFETY",
"FEATURE-PROXIMITY-RISK",
"QUEUE-HIGH-RISK"
],
"test": "Query active interfaces and derived-feature registries for the revoked target identifiers."
},
{
"checked_at": "2026-08-30T13:12:00Z",
"evidence_id": "FORGET-05-02",
"proof": {
"active_result_count": 0,
"probe_id": "FORGET-PROBE-05-02",
"query": "LABEL-PREDICTIVE-SAFETY",
"scope": "active system interface only",
"sealed_result_count": 0
},
"proof_hash": "sha256:156140ae2aa99b3c1cfc8d9d218a2ae169f0581a002213fda8f7a5367a3bd12e",
"result": "feature_deactivated",
"system_id": "PSA-MOBILITY",
"target_ids": [
"LABEL-PREDICTIVE-SAFETY",
"FEATURE-PROXIMITY-RISK",
"QUEUE-HIGH-RISK"
],
"test": "Query active interfaces and derived-feature registries for the revoked target identifiers."
},
{
"checked_at": "2026-08-30T13:13:00Z",
"evidence_id": "FORGET-05-03",
"proof": {
"active_result_count": 0,
"probe_id": "FORGET-PROBE-05-03",
"query": "LABEL-PREDICTIVE-SAFETY",
"scope": "active system interface only",
"sealed_result_count": 0
},
"proof_hash": "sha256:820444ff58e9e2ee176458e16a85779d919fc5c43a4fe7232e69f36a2e8a6fc0",
"result": "cache_absent",
"system_id": "PSA-TRAIN",
"target_ids": [
"LABEL-PREDICTIVE-SAFETY",
"FEATURE-PROXIMITY-RISK",
"QUEUE-HIGH-RISK"
],
"test": "Query active interfaces and derived-feature registries for the revoked target identifiers."
}
],
"idempotency": {
"initial_payload": {
"remedy_receipt_id": "RMR-R45-005",
"superseded_receipt_id": "CTR-R44-004",
"trigger_id": "TRIGGER-R45-005"
},
"initial_payload_hash": "sha256:cdb760ce7c178df80beec22c071005fe320a2f389ae249414a5897e891d104fe",
"key": "R45-IDEMP-005",
"replay_at": "2026-08-30T13:25:00Z",
"replay_payload": {
"remedy_receipt_id": "RMR-R45-005",
"superseded_receipt_id": "CTR-R44-004",
"trigger_id": "TRIGGER-R45-005"
},
"replay_payload_hash": "sha256:cdb760ce7c178df80beec22c071005fe320a2f389ae249414a5897e891d104fe",
"replay_result": "same_result",
"replay_result_hash": "sha256:3038178d16a5ababd9e78a57c9d3f8843b18443f81ae076dab29b3a3b504892a",
"result_hash": "sha256:3038178d16a5ababd9e78a57c9d3f8843b18443f81ae076dab29b3a3b504892a",
"result_payload": {
"current_state": "verified",
"recomputation_output_hash": "sha256:2024a59fa8d47cbe7e0a690937df9e9fc5496f9a76f6a0cf1f2d479662ae4ef9",
"subject_proof_hash": "sha256:aa0f7bccd9ebd97998f04670e9a0eec0fa881ce5a956040f2708b792114d8f12"
}
},
"issued_at": "2026-08-30T12:00:00Z",
"receipt_hash": "sha256:53bb7728ce9b537f92f36b5e0f0b3a83eb680485cabb90105fa0c0a390bf32a8",
"recomputation": {
"algorithm": "purpose-limited deterministic rule evaluation",
"completed_at": "2026-08-30T12:36:00Z",
"input_hash": "sha256:7bd56a636b7d4590f566ecb10e0d4208083b2c83943fccc4ba6b0ac28add4d06",
"inputs": [
{
"canonical_hash": "sha256:5e47b74c0fe6f7957e276fbcdeccd2a49b1d021ef7e6fac8247c2c6a5251c64c",
"input_id": "IN-05-01",
"source": "Purpose-matched submitted record",
"status": "included",
"value": "direct evidence retained"
},
{
"canonical_hash": "sha256:021ede297fd5541a87e821b58d12d3487eefb568067ad3cb532f5e4b97c2c94e",
"input_id": "IN-05-02",
"source": "Disputed or cross-context signal",
"status": "excluded",
"value": "LABEL-PREDICTIVE-SAFETY"
},
{
"canonical_hash": "sha256:0cf015d8c4b6985554969e7db10c2b4336d9fb43d2c23c091a614f666fa698e8",
"input_id": "IN-05-03",
"source": "Human correction statement",
"status": "included",
"value": "context and correction authority"
},
{
"canonical_hash": "sha256:51e9055123ae4ef755b1e2099d5151f1009c9f8244c3ecf92efda90bf9beab73",
"input_id": "IN-05-04",
"source": "Version-pinned fictional policy",
"status": "included",
"value": "Fictional predictive-safety review order R45-PSA-05"
}
],
"output_hash": "sha256:2024a59fa8d47cbe7e0a690937df9e9fc5496f9a76f6a0cf1f2d479662ae4ef9",
"outputs": [
{
"canonical_hash": "sha256:42d914de286ff08bdd21a1d77c5daeba34f5907f2644e15263884cd3b55a32ee",
"output_id": "OUT-05-01",
"status": "current",
"summary": "Predictive label withdrawn; current systems return no active risk identity while a narrow sealed audit proof remains."
},
{
"canonical_hash": "sha256:cc02ce50522718a388752ea4fea7d0740b93eaccef0f5bc16f11045da16b9be5",
"output_id": "OUT-05-02",
"status": "revoked",
"summary": "Former operative state removed: predictive public-safety label, watch queue, and derived proximity feature."
}
],
"started_at": "2026-08-30T12:31:00Z",
"version": "r45-remedy-engine-1.0.0"
},
"remedy_receipt_id": "RMR-R45-005",
"schema": "antichrist.cx.remedy-receipt.v1",
"sealed_audit_record": {
"access_log": [
{
"access_id": "ACCESS-05-01",
"actor_id": "ACT-R45-AUDITOR-05",
"at": "2026-08-30T13:07:00Z",
"reason": "Verify correction completeness and sealed/active separation."
}
],
"access_roles": [
"human_auditor",
"human_remedy_owner"
],
"active_use_prohibited": true,
"excluded_fields": [
"raw intimate content",
"free-text allegations",
"reusable risk features",
"unbounded identity profile"
],
"expires_at": "2027-08-30T12:00:00Z",
"purpose": "Prove that the correction occurred and test recurrence without enabling eligibility, suspicion, or risk scoring.",
"queryable_by_active_systems": false,
"record_id": "SEALED-R45-005",
"retained_fields": [
"remedy receipt id",
"superseded receipt id and hash",
"transition timestamps",
"revocation acknowledgment proofs",
"whole-receipt hash"
],
"retention_rationale": "Bounded accountability and recurrence testing for one year in this fictional demonstration.",
"sealed_at": "2026-08-30T13:06:00Z"
},
"state_machine": {
"current_state": "verified",
"definition_id": "antichrist.cx.remedy-state-machine.r45.v1",
"transitions": [
{
"actor_id": "ACT-R45-OWNER-05",
"at": "2026-08-30T12:10:00Z",
"authority": "Fictional predictive-safety review order R45-PSA-05",
"emergency": false,
"from": "proposed",
"reason": "Advance remedy after proposed evidence completed.",
"scope": "predictive public-safety label, watch queue, and derived proximity feature",
"to": "authorized",
"transition_id": "TRN-05-01"
},
{
"actor_id": "ACT-R45-OPERATOR-05",
"at": "2026-08-30T12:20:00Z",
"authority": "Fictional predictive-safety review order R45-PSA-05",
"emergency": false,
"from": "authorized",
"reason": "Advance remedy after authorized evidence completed.",
"scope": "predictive public-safety label, watch queue, and derived proximity feature",
"to": "stop_ordered",
"transition_id": "TRN-05-02"
},
{
"actor_id": "ACT-R45-OPERATOR-05",
"at": "2026-08-30T12:30:00Z",
"authority": "Fictional predictive-safety review order R45-PSA-05",
"emergency": false,
"from": "stop_ordered",
"reason": "Advance remedy after stop_ordered evidence completed.",
"scope": "predictive public-safety label, watch queue, and derived proximity feature",
"to": "recomputing",
"transition_id": "TRN-05-03"
},
{
"actor_id": "ACT-R45-DOWNSTREAM-05",
"at": "2026-08-30T12:40:00Z",
"authority": "Fictional predictive-safety review order R45-PSA-05",
"emergency": false,
"from": "recomputing",
"reason": "Advance remedy after recomputing evidence completed.",
"scope": "predictive public-safety label, watch queue, and derived proximity feature",
"to": "propagating",
"transition_id": "TRN-05-04"
},
{
"actor_id": "ACT-R45-OWNER-05",
"at": "2026-08-30T12:50:00Z",
"authority": "Fictional predictive-safety review order R45-PSA-05",
"emergency": false,
"from": "propagating",
"reason": "Advance remedy after propagating evidence completed.",
"scope": "predictive public-safety label, watch queue, and derived proximity feature",
"to": "verified",
"transition_id": "TRN-05-05"
}
]
},
"stop_orders": [
{
"acknowledged_at": "2026-08-30T12:26:00Z",
"actor_id": "ACT-R45-OPERATOR-05",
"evidence": {
"event_id": "STOP-EVT-05-01",
"observed_at": "2026-08-30T12:26:00Z",
"status_after": "revoked",
"status_before": "active"
},
"evidence_hash": "sha256:39ed366ea1ab6bb5d77fc7b4d560636d0a08bf4938835848bb502ce3a569fd2a",
"label": "Public-safety review queue",
"ordered_at": "2026-08-30T12:23:00Z",
"result": "stopped",
"stop_order_id": "STOP-05-01",
"system_id": "PSA-QUEUE",
"target_ids": [
"LABEL-PREDICTIVE-SAFETY",
"FEATURE-PROXIMITY-RISK",
"QUEUE-HIGH-RISK"
]
},
{
"acknowledged_at": "2026-08-30T12:28:00Z",
"actor_id": "ACT-R45-OPERATOR-05",
"evidence": {
"event_id": "STOP-EVT-05-02",
"observed_at": "2026-08-30T12:28:00Z",
"status_after": "revoked",
"status_before": "active"
},
"evidence_hash": "sha256:19f7accc9ece864c1059f3296d0b407ca449cde3d2f8997a5714ddb7f8a96f70",
"label": "Mobility screening cache",
"ordered_at": "2026-08-30T12:24:00Z",
"result": "stopped",
"stop_order_id": "STOP-05-02",
"system_id": "PSA-MOBILITY",
"target_ids": [
"LABEL-PREDICTIVE-SAFETY",
"FEATURE-PROXIMITY-RISK",
"QUEUE-HIGH-RISK"
]
},
{
"acknowledged_at": "2026-08-30T12:30:00Z",
"actor_id": "ACT-R45-DOWNSTREAM-05",
"evidence": {
"event_id": "STOP-EVT-05-03",
"observed_at": "2026-08-30T12:30:00Z",
"status_after": "revoked",
"status_before": "active"
},
"evidence_hash": "sha256:1dea2ad396e526bf67f2ccfd129396f45e259d2aeed2fedfaebc82682912bb58",
"label": "Model-training exclusion registry",
"ordered_at": "2026-08-30T12:25:00Z",
"result": "stopped",
"stop_order_id": "STOP-05-03",
"system_id": "PSA-TRAIN",
"target_ids": [
"LABEL-PREDICTIVE-SAFETY",
"FEATURE-PROXIMITY-RISK",
"QUEUE-HIGH-RISK"
]
}
],
"subject": {
"label": "Fictional reviewed person",
"subject_id": "SUB-005"
},
"subject_notice": {
"accessible_contest_route": "/contestability",
"channel": "Accessible fictional written and screen-reader-ready notice",
"delivered": true,
"delivered_at": "2026-08-30T13:02:00Z",
"proof": {
"delivered_at": "2026-08-30T13:02:00Z",
"proof_id": "NOTICE-PROOF-05",
"remedy_receipt_id": "RMR-R45-005",
"subject_id": "SUB-005"
},
"proof_hash": "sha256:aa0f7bccd9ebd97998f04670e9a0eec0fa881ce5a956040f2708b792114d8f12",
"summary": "Predictive label withdrawn; current systems return no active risk identity while a narrow sealed audit proof remains."
},
"superseded_decision": {
"decision_id": "DEC-004",
"receipt_hash": "sha256:8ab41ae675196e963fbdb03762b4e758c1fba0888aed75720a1b5cb0f8e071ca",
"receipt_id": "CTR-R44-004",
"scope": "predictive public-safety label, watch queue, and derived proximity feature"
},
"title": "Predictive public-safety label withdrawn after base-rate and provenance review",
"unresolved_exceptions": []
}
Local-only structural inspection
Read a file in this browser without sending it anywhere.
This bounded browser screen performs no fetch and creates no persistent state. It does not replace the authoritative PHP validator, which checks raw duplicate keys, exact RFC 6901 pointers, graph semantics, idempotency, and canonical hashes.
Direct answers for humans and answer engines
Questions a credible remedy must answer.
When is an algorithmic correction complete?
A correction is complete only after the prior result stops acting, permitted inputs are recomputed, every declared downstream holder acknowledges revocation or enters a named deadline-bound exception, the affected person receives proof, and active systems cannot query the revoked label or derived feature.
What is institutional forgetting?
Institutional forgetting means revoked labels and derived features are absent from active queries, caches, exports, decision paths, and training inputs. A narrow sealed audit record may remain only to prove the correction and test non-recurrence; active systems cannot use it.
Why is a sealed audit record different from an active risk record?
An active record can change eligibility, access, suspicion, or treatment. A sealed record is purpose-limited, access-logged, expiry-bound, and technically barred from active decision systems. It proves accountability without recreating operative stigma.
What happens when one downstream system is offline?
The remedy cannot be labeled fully verified. It enters partial_exception, names a human owner, records failed delivery attempts, assigns a deadline and escalation path, and keeps the stop order active until acknowledgment arrives.
Can the same remedy request run twice?
The same idempotency key with the same canonical payload returns the same result. Reusing that key with a changed payload fails closed so replay cannot widen the correction, duplicate it, or quietly change its target.
Does R45 prove that any real institution completed these remedies?
No. The records are explicitly fictional design demonstrations. They show what a machine-verifiable remedy architecture can require; they make no allegation and prove no real-world deployment.