Evidence schema by format
Offsets use baseline proof, receipt or audit links, match ratio, surplus rule, and destination checks. Pledge swaps use logs, attestations, timestamps, and agreed check-ins.
Validator scope
What reviewers are allowed to certify.
Reviewers certify narrow evidence claims, not broad moral worth, legal enforceability, tax treatment, escrow status, or final real-world impact.
One proof, one claim
The same receipt, audit, or public log should not create multiple verified-completion claims unless the record explicitly explains why reuse is valid.
Conflict rules
Reviewers should not validate offers where they are a party, beneficiary, close counterparty, or institutional sponsor.
Appeal and challenge lane
Participants can challenge a completion state, duplicate proof, coercive baseline, or factual error before a badge becomes durable.
Reviewer quality metrics
Track turnaround time, challenge rate, appeal overturn rate, duplicate-proof misses, and unresolved dispute share.
Public transparency note
Publish real counts only: submitted records, reviewed completions, blocked proposals, disputes, and appeal outcomes.
Public validator
The core Moral Trade protocol now has a public check surface.
The validator publishes required proposal fields, status values, factor codes, evidence schemas, and provenance objects so the core feature is inspectable like MPGF.
Core profile
moral-trade-core-v0.1.6-2026-06
11 check(s), 0 blocker(s), status pass.
Status taxonomy
Every proof claim should have a visible state.
State 1
Draft
Terms are being written and are not visible as live liquidity.
No reviewer action yet.
State 2
Submitted
The participant has stated the action, reciprocal request, baseline, duration, exit condition, and evidence rule.
Completeness and safety screen.
State 3
Needs evidence
The claim depends on receipts, public logs, third-party records, or prior-donation proof.
Request missing proof or keep the record paused.
State 4
Challenge window
Counterparties or reviewers can flag duplicate proof, coercive framing, or factual gaps.
Hold completion until the challenge lane closes.
State 5
Completion reviewed
Evidence supports the specific claim being displayed, without implying legal, tax, custody, or outcome guarantees.
Publish only the verified claim and reviewer confidence band.
State 6
Disputed or unresolved
The record remains visible as a problem state, not a completed trade.
Publish a short reasoning summary when disclosure is safe.
Reviewer roles
Manual review has named responsibilities.
Each role certifies a narrow operational question so users can tell whether a record is complete, evidenced, appealed, or still unresolved.
Intake reviewer
Checks completeness, prohibited categories, baseline clarity, evidence method, and whether the proposal belongs in the public launch wedge.
Publishes: A submitted, paused, or blocked intake state with the narrow reason safe to disclose.
Evidence reviewer
Checks receipts, public logs, attestations, provider records, proof uniqueness, and challenge-window readiness.
Publishes: A needs-evidence, challenge-window, completion-reviewed, or disputed/unresolved state.
Appeal reviewer
Reviews challenged decisions when new evidence, conflict concerns, duplicate proof, or coercive-baseline claims are raised.
Publishes: A short reasoning summary, outcome, and any changed evidence state when disclosure is safe.
Operating targets
Review needs SLAs, conflict rules, and appeals.
These targets turn the trust problem from a site-wide disclaimer into a visible operating promise that can be audited over time.
Before assignment
Conflict rule
Reviewers should not validate records where they are a party, beneficiary, funder, close counterparty, or institutional sponsor.
2 business days
Intake SLA
New public offers should receive a completeness and safety screen before they are treated as live marketplace supply.
5 business days
Evidence SLA
Submitted evidence should be marked as missing, in challenge window, review-cleared, or unresolved within a visible timeframe.
7 business days
Appeal lane
Appeals should preserve the original decision, challenger claim, reviewer response, and final state in the audit log.
Challenge and appeal contract
Appeals are scoped to reviewable claims, not general moral disagreement.
Challenges can target a specific claim, evidence row, baseline concern, disclosure decision, externality trigger, completion state, or policy flag. They do not mutate live state by themselves; they route a human review lane with redacted provenance.
Appeal contract moral-trade-challenge-appeal-v0.2
pass
pass9 check(s), 0 blocker(s), 8 trigger(s).
Reviewable subjects
claim, evidence row, baseline concern, disclosure decision, externality trigger, completion state, policy flag.
Standing categories
participant, counterparty, affected party, reviewer, admin safety, external verifier.
Appeal triggers
duplicate proof, coercive baseline, wrong scope evidence, material factual error, privacy disclosure error, externality remedy gap, reviewer conflict, policy misapplied.
Allowed outcomes
uphold decision, request evidence, route human review, open challenge window, block reliance, record remedy, close unresolved, correct record.
Requested outcomes are advisory and must match the appeal trigger before reviewer routing.
Reviewer quality
Trust metrics should be published before trust badges scale.
The pilot should track reviewer performance and failure modes before expanding reputation or paid-action volume.
Median intake time
Median evidence-review time
Challenge rate
Appeal overturn rate
Duplicate-proof misses
Unresolved dispute share
Trust ladder
Badges are transaction-linked, never decorative.
Identity verified
Show this badge only when the underlying record and review scope support it.
Organization verified
Show this badge only when the underlying record and review scope support it.
Provider payment verified
Only provider-linked receipts or webhooks should create this badge.
Completion reviewed
Show this badge only when the underlying record and review scope support it.
Repeat counterparty
Show this badge only when the underlying record and review scope support it.
Governance
Centralized operator, published rules, independent review.
The next operating model is founder-led moderation with a public rulebook, an appeal path, external reviewer panel, and real transparency numbers.
Rulebook first
No threats, coercive baselines, hidden platform fees, unsupported jurisdictions, or campaign-contribution offsets.
Appeals and audits
Hard cases should leave an internal audit log and, when safe, a short publishable reasoning summary.
Portable later
The pilot stays centralized for safety and compliance while preserving exportable records for future interoperability.