Lists approval requests across the agents of the caller's Applications, newest first
Read-only companion to the agent-scoped `GET .../agents/{agent}/pendingApprovals`. Each item carries its `agentId`, so inbox rows can route their respond/cancel actions to the agent-scoped endpoints (which enforce per-agent Admin access). Naming an Application — or an `agentId` in an Application — the caller cannot open is a 403. The `networkId` filter mirrors the one the escalation queue and access-request list already take, so the Inbox can scope all three of its folders server-side instead of post-filtering a tenant-wide page (which silently under-fills the page it is filtering).
Authorization
Bearer JWT Authorization header using the Bearer scheme. Enter 'Bearer' [space] and then your token.
In: header
Path Parameters
The tenant identifier
Query Parameters
Filter by lifecycle state. Defaults to Fruxon.Model.HumanApprovalRequests.HumanApprovalStatus.Pending.
"UNSPECIFIED" | "PENDING" | "RESPONDED" | "CANCELLED" | "EXPIRED"Optional Application filter. Asserted against the caller's access.
uuidOptional filter to a single agent. Asserted against its owning Application.
The maximum number of items to return per page
int32int32Opaque cursor token from an external system (e.g. RAG API). When set, M:Fruxon.Common.Collections.PageToken.AsString returns this value directly instead of encoding Fruxon.Common.Collections.PageToken.Skip.
Response Body
curl -X GET "https://api.fruxon.com/v1/tenants/string/pendingApprovals"{
"items": [
{
"id": "00000000-0000-0000-0000-000000000000",
"agentId": "string",
"agentRevision": 0,
"flow": "string",
"sessionId": "string",
"taskQueueId": 0,
"executionRecordId": "00000000-0000-0000-0000-000000000000",
"gateType": "TOOL_FLOW_STEP",
"stepIdentifier": "string",
"toolCallId": "string",
"toolKey": "string",
"outboundMessage": "string",
"proposedParameters": {},
"authorizedApprovers": [
"string"
],
"deliveryRoute": {
"id": "00000000-0000-0000-0000-000000000000",
"connectorId": "00000000-0000-0000-0000-000000000000",
"provider": "string",
"integrationConfigId": null,
"environmentSlug": "string",
"parameters": null
},
"deliveryMessageId": "string",
"correlationToken": "string",
"status": "PENDING",
"expiresAt": 0,
"respondedAt": 0,
"respondedBy": "string",
"decision": "APPROVE",
"responseText": "string",
"createdAt": 0,
"modifiedAt": 0
}
],
"nextPageToken": "string",
"totalCount": 0
}{
"type": "string",
"title": "string",
"status": 0,
"detail": "string",
"instance": "string",
"property1": null,
"property2": null
}{
"type": "string",
"title": "string",
"status": 0,
"detail": "string",
"instance": "string",
"property1": null,
"property2": null
}Gets a single approval request by id.
Returns the full record for one human-approval gate, including its current lifecycle state, the tool call payload awaiting approval, the parent execution id, and (if already resolved) the responder identity, decision, and decision text. Returns 404 if the approval id does not belong to the agent in the route — approval ids are agent-scoped and cross-agent access is not permitted even with tenant-admin credentials.
Records the human decision and reopens the parent task for worker resume
On approve, the parent execution is requeued from the parked step and the tool call proceeds with its original arguments; on reject, the rejection text is fed back into the flow as the tool result so the agent can react to (or surface) the operator's reasoning. The responder identity is captured from the caller's auth context, not the request body, so audit trails reflect the real user. Requires Admin role on the agent to prevent arbitrary viewers from authorizing privileged tool calls; to abandon a request without resuming, use `POST .../pendingApprovals/{pendingApproval}:cancel`.