Policy fields

All 61 fields a policy rule can test, with their types and operators.

A policy rule tests fields of the evaluation context. The list is closed: an unknown field is a validation error at authoring time, not a silent undefined at decision time. That is what makes evaluation total — every rule either matches or does not, and there is no third outcome.

Amounts are inputs

amount is in minor currency units and exists purely as a policy input and as recorded decision context. Veefore evaluates it and records it. It does not hold, authorise, capture or move anything.

The same list is served at GET /v1/policies/builder-schema, which is what the visual Policy Builder renders from.

#Agent

Field	Type	Operators	Meaning
`agent.id`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Immutable agent identifier.
`agent.name`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Display name. Metadata only — never a security identity.
`agent.status`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Lifecycle state.
`agent.risk_level`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Standing risk rating of the agent.
`agent.verification_status`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Identity verification state.
`agent.protocols`	`string[]`	`contains` `not_contains` `contains_any` `contains_all` `eq` `neq` `exists` `not_exists`	Protocols the agent is registered for.
`agent.capabilities`	`string[]`	`contains` `not_contains` `contains_any` `contains_all` `eq` `neq` `exists` `not_exists`	Capabilities currently granted to the agent.
`agent.tags`	`string[]`	`contains` `not_contains` `contains_any` `contains_all` `eq` `neq` `exists` `not_exists`	Operator-assigned labels.

agent.status accepts: PROVISIONING, ACTIVE, PAUSED, SUSPENDED, REVOKED, EXPIRED, DECOMMISSIONED.

agent.risk_level accepts: LOW, MEDIUM, HIGH, CRITICAL.

agent.verification_status accepts: UNVERIFIED, PENDING, VERIFIED, FAILED.

agent.protocols accepts: HTTP, MCP, A2A.

#Tenancy

Field	Type	Operators	Meaning
`organization.id`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Organization identifier.
`workspace.id`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Workspace identifier.
`environment.id`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Environment identifier.
`environment.type`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Deployment tier.
`environment.name`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Environment display name.

environment.type accepts: DEVELOPMENT, STAGING, PRODUCTION.

#Principal

Field	Type	Operators	Meaning
`principal.type`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Who initiated the request.
`principal.id`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Identifier of the initiating principal.

principal.type accepts: USER, AGENT, SYSTEM, API_KEY.

#Tool

Field	Type	Operators	Meaning
`tool.id`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Registered tool identifier.
`tool.key`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Tool key, e.g. refund_customer.
`tool.kind`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Execution kind.
`tool.protocol`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Transport protocol.
`tool.risk_level`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Risk rating assigned to the tool.
`tool.dangerous`	`boolean`	`is_true` `is_false` `eq` `neq`	Tool is flagged as destructive or high impact.

tool.kind accepts: HTTP, MCP, A2A, DATABASE, WEBHOOK, SIMULATED, CLIENT.

tool.protocol accepts: HTTP, MCP, A2A.

tool.risk_level accepts: LOW, MEDIUM, HIGH, CRITICAL.

#Action

Field	Type	Operators	Meaning
`capability`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Capability required by the request, e.g. refund.create.
`action`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Logical operation being performed.

#Resource

Field	Type	Operators	Meaning
`resource.key`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Target resource identifier.
`resource.type`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Resource category, e.g. database.
`resource.sensitivity`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Data classification.
`resource.attributes.<path>`	`object`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists` `eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Arbitrary resource attribute by path.

resource.sensitivity accepts: PUBLIC, INTERNAL, CONFIDENTIAL, RESTRICTED.

#Request

Field	Type	Operators	Meaning
`amount`	`amount`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Requested amount converted to the organization base currency, in minor units (e.g. 1000000 = ₹10,000.00 when the base is INR). Thresholds written here are in the base currency. A request whose currency cannot be converted is held for approval before policy amounts are relied on. Veefore never moves money.
`currency`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Original ISO 4217 currency of the request, e.g. USD. Not the base currency.
`original_amount`	`amount`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Amount exactly as requested, in minor units of the original currency (no conversion). Combine with a currency condition when comparing it.
`original_currency`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	ISO 4217 currency the request was made in (same as "Currency").
`destination.host`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Outbound hostname the tool will contact.
`destination.risk_level`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Risk rating of the destination.
`parameters.<path>`	`object`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists` `eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists` `is_true` `is_false` `eq` `neq`	Request parameter by dotted path.

destination.risk_level accepts: LOW, MEDIUM, HIGH, CRITICAL.

#Risk

Field	Type	Operators	Meaning
`risk.level`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Computed risk level for this request.
`risk.score`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Weighted risk score, 0-100.
`risk.signals`	`string[]`	`contains` `not_contains` `contains_any` `contains_all` `eq` `neq` `exists` `not_exists`	Signal types that fired for this request.

risk.level accepts: LOW, MEDIUM, HIGH, CRITICAL.

#Trust

Field	Type	Operators	Meaning
`trust.identity_verified`	`boolean`	`is_true` `is_false` `eq` `neq`	Valid identity verification evidence exists.
`trust.organization_verified`	`boolean`	`is_true` `is_false` `eq` `neq`	Valid organization verification evidence exists.
`trust.security_attestation_valid`	`boolean`	`is_true` `is_false` `eq` `neq`	Unexpired security attestation evidence exists.
`trust.protocol_compliance_valid`	`boolean`	`is_true` `is_false` `eq` `neq`	Unexpired protocol compliance evidence exists.
`trust.active_incidents`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Open incidents involving this agent.
`trust.expired_evidence_count`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Trust evidence records that have expired.

#Authority

Field	Type	Operators	Meaning
`authority.max_amount`	`amount`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Maximum amount permitted by delegated authority, converted to the organization base currency (minor units).
`authority.capabilities`	`string[]`	`contains` `not_contains` `contains_any` `contains_all` `eq` `neq` `exists` `not_exists`	Capabilities permitted by delegated authority.
`authority.depth`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Delegation depth backing the authority.
`delegation.depth`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Depth of the delegation used.
`delegation.chain_length`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Number of delegations in the chain.
`delegation.valid`	`boolean`	`is_true` `is_false` `eq` `neq`	Delegation chain resolved successfully.

#Context

Field	Type	Operators	Meaning
`time.hour_utc`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Hour of day, 0-23, in UTC.
`time.day_of_week`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	0 = Sunday through 6 = Saturday, UTC.
`time.iso`	`string`	`eq` `neq` `in` `not_in` `contains` `not_contains` `starts_with` `ends_with` `matches` `exists` `not_exists`	Request timestamp, ISO-8601 UTC.
`velocity.actions_last_minute`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Agent action count in the trailing 60 seconds.
`velocity.actions_last_hour`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Agent action count in the trailing hour.
`velocity.actions_last_day`	`number`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Agent action count in the trailing 24 hours.
`velocity.amount_last_day`	`amount`	`eq` `neq` `gt` `gte` `lt` `lte` `between` `in` `not_in` `exists` `not_exists`	Sum of amounts in the trailing 24 hours, in the organization base currency.
`history.tool_used_before`	`boolean`	`is_true` `is_false` `eq` `neq`	Agent has previously used this tool.
`history.destination_used_before`	`boolean`	`is_true` `is_false` `eq` `neq`	Agent has previously contacted this destination.
`history.resource_used_before`	`boolean`	`is_true` `is_false` `eq` `neq`	Agent has previously touched this resource.

#Approval

Field	Type	Operators	Meaning
`approval.present`	`boolean`	`is_true` `is_false` `eq` `neq`	Request carries an approval reference.
`approval.valid`	`boolean`	`is_true` `is_false` `eq` `neq`	Presented approval is granted and bound to this request.