admin
Cross-tenant Run ledger
List agent runs across workspaces, with filters for workspace, person, conversation, status, trigger, source, time, cost, and silent outcomes.
Returns items and paging information. Each item includes recorded tokens, cost, provenance, and outcome flags.
get/v1/admin/runs
Request
- Base URL: https://api.patricia.app (the document declares no server; this is the origin it was published from)
- URL: https://api.patricia.app/v1/admin/runs
- Auth: HTTP bearer
Query parameters
team_idstring uuid nullable
actor_idstring uuid nullable
conversation_idstring uuid nullable
parent_run_idstring uuid nullable
status_code'queued' | 'running' | 'waiting_approval' | 'suspended' | 'completed' | 'failed' | 'cancelled' | 'denied_autonomous'
trigger_code'mention' | 'dm' | 'message' | 'proactive' | 'cron' | 'event' | 'heartbeat' | 'discovery' | 'onboarding' | 'manual' | 'spawn'
source_kind'task_cron' | 'event_trigger' | 'scheduled_program' | 'cued_proactive' | 'sweep_proactive' | 'onboarding' | 'manual' | 'spawn'
What kind of durable thing caused a Class-A run, for AgentRun.source_kind.
The source is the JOB, PROGRAM, or EVENT, deliberately not the run and not the tool family. A daily cron is ONE source that fires ninety times, and the question Phase 1 has to answer ("how much would this source actually be asking a human?") is unanswerable if each fire looks like an unrelated run.
source_refstring nullable
modelstring nullable
cost_micro_usd_mininteger nullable
cost_micro_usd_maxinteger nullable
duration_ms_mininteger nullable
duration_ms_maxinteger nullable
started_afterstring date-time nullable
started_beforestring date-time nullable
created_afterstring date-time nullable
created_beforestring date-time nullable
truncatedboolean nullable
no_resultboolean nullable
silentboolean nullable
limitinteger
sort_by'created_at:desc' | 'created_at:asc'
pqstring nullable
from_prevboolean
Response
Successful Response
Changes
No recorded changes to this operation across all 1 revision of this API.