OrchestrationAgentResource

The prompt that tells Resia how to read the run's outcome into the analysis result.

A JSON Schema (draft 2020-12) the run's analysis result is shaped by. Use a JSON Schema object, not the legacy fields format.

The latest local clock time a run may place a call, as HH:MM:SS. It must be after calling_window_start_local_time.

The earliest local clock time a run may place a call, as HH:MM:SS. The window has no time zone of its own; each run's inputs carry the time zone the window is read in. The start must be before the end.

When the stable Orchestration Agent was created.

The stable Orchestration Agent identifier.

A JSON Schema (draft 2020-12) the inputs of each run must satisfy. Use a JSON Schema object, not the legacy fields format.

The instructions for each run. Use $name or ${name} to refer to a top-level run input property. Declare every placeholder in input_schema.properties. Resia supplies validated run inputs to the agent as a separate JSON object. Resia does not replace the placeholder text. Nested property paths are not supported.

Whether the agent has an approved production configuration and is not archived.

The most calls one run of this Orchestration Agent may place. Zero means the run places no calls, and it then receives no calling ability at all. A run that reaches this limit is refused further calls.

The most one run of this Orchestration Agent may spend, in whole cents. A run that reaches this limit is refused further spend.

The Orchestration Agent's name. It must be unique within your organization.

Whether the latest submitted configuration is approved or still pending review.

When the latest configuration was submitted.

When the current active configuration was approved, or null when none is active.

When the agent was archived, or null.

One line saying what this Orchestration Agent is for. Optional; a name is the fallback.

The most text messages one run of this Orchestration Agent may send. Zero (the default) means the run sends none, and it then receives no messaging ability.

The stages a run of this Orchestration Agent may report, each a name and a description. During a run the agent records the stage it has reached, using a name from this list; GET the run to read the current stage. Names must be unique. Null (the default) means the agent reports no milestone.

HTTPS URL Resia POSTs an orchestration_run.ended event to when an Orchestration Run reaches succeeded, failed, or cancelled. The event contains the same OrchestrationRunResponse that GET /v1/orchestration-runs/{orchestration_run_id} returns. Resia makes one immediate attempt when practical and retries a failed delivery on the call.ended schedule. Any 2xx response is success. The orchestration_run_id stays unchanged across retries, so treat a repeated orchestration_run_id as a duplicate. Treat the complete URL as a credential. Null clears the webhook and stops pending deliveries.

Open key/value labels on this record. Set any keys and any values. Filter a list by one key and value pair. Resia never interprets a tag. Do not put names, phone numbers, or any personal, health, credential, or secret information in a tag: a tag travels in URLs when used as a filter, and URLs appear in ordinary infrastructure request logs.

The names of your organization's tools that runs of this version may call. Null or an empty list names none.

The Call Agents runs of this Orchestration Agent may dial, each named by its identifier. Every identifier must name a Call Agent your organization owns and has not archived. An empty list (the default) means the orchestration places no calls.

The Chat Agents runs of this Orchestration Agent may start a chat with, each named by its identifier. Every identifier must name a Chat Agent your organization owns and has not archived. An empty list (the default) means the orchestration starts no chats. A run also needs max_messages above zero to start a chat.