Domain Explorer
Each domain defines a specific trust question with its own evaluation policy, outcomes, and consumption model.
What is a Domain?
A DomainSpec defines the semantic boundary for judgment. Each domain answers a specific trust question with its own:
- Evidence Policy - What data is admissible and how it's selected
- Evaluation Spec - What features are required and how they map to outcomes
- Model Policy - Whether evaluation uses LLM consensus or deterministic rules
- Equivalence Profile - When two judgments are considered equivalent
- Revision Policy - How judgments evolve over time
The key insight: not one generic reputation score, but domain-specific trust signals.
Protocol Objects
DomainSpec
Semantic boundary and policy
CaseObject
Reproducible evidence artifact
AssessmentArtifact
Evaluation record and payload
JudgmentObject
Compact trust primitive
Live Domains
Counterparty Trust
counterparty_trust.base_trade_v1
Judge whether a wallet is safe to trade with.
Possible Outcomes
SAFECAUTIONUNSAFEINSUFFICIENT_DATA
Claim Type
counterparty_trade_readiness
Window Policy
Moderolling
Default Window5,000 blocks
Freshness50,000 blocks
Required Features
tx_countfailed_tx_countunique_counterpartiesunbounded_approval_countflagged_interaction_count
Supported Networks
base-sepoliabase-mainnet
Primary Consumer Question
Should this wallet be trusted as a trading counterparty in this observation window?
Protocol Safety
protocol_safety.base_erc20_permission_v1
Judge whether ERC-20 approval posture is safe.
Possible Outcomes
SAFECAUTIONUNSAFEINSUFFICIENT_DATA
Claim Type
erc20_permission_safety
Window Policy
Moderolling
Default Window100 blocks
Freshness50,000 blocks
Required Features
approval_breadthspender_specific_riskunbounded_approvalsfailed_approval_actions
Supported Networks
base-sepoliabase-mainnet
Primary Consumer Question
Is this ERC-20 approval posture safe for the wallet and spender in this observation window?
Hot-Path Consumption Rules
For all domains, the consumption flow follows this pattern:
SAFE
Allow when fresh
CAUTION
Review or tighten policy
UNSAFE
Deny the interaction
INSUFFICIENT_DATA
Refresh or collect more evidence
STALE
Refresh before execution-grade trust
NO_JUDGMENT
Collect and submit a case first