Automation MCP response

MCP tool response returned by the Search automations, Create automation, and Update automation MCP tools.

Returned by: Search automations MCP tool; Create automation MCP tool; Update automation MCP tool
Response type: Structured automation record

Response Shape

The returned automation record can include identity fields, the current prompt and reasoning level, completion email settings, ordered start nodes, run-history snapshots, and status metadata.

Field	Type	Notes
`govtribe_id`	string	GovTribe ID for the automation. Pass this value into `workflow_id` for update, delete, or test-run calls.
`govtribe_type`	string	GovTribe record type for the automation row.
`govtribe_url`	string	GovTribe URL for the automation detail page.
`name`	string	Automation name.
`prompt`	string	Prompt GovTribe AI runs whenever any start node fires.
`reasoning_level`	string	Reasoning used for automation runs. Values are `none`, `low`, `medium`, and `high`.
`on_completion_notify_user`	string	Completion email setting. Values are `always`, `conditionally`, and `never`.
`completion_notification_instructions`	string or null	Condition and email-writing guidance for the completion email.
`start_nodes`	object array	Ordered start triggers. Each item contains a stable ID and a round-trippable public definition.
`total_triggers_count`	integer	Complete number of start triggers, including any omitted from a truncated search response.
`trigger_details`	object or null	Compatibility snapshot of the first ordered start node. Use `start_nodes` for complete trigger inspection and edits.
`state`	string	Current automation state such as `published`, `paused`, or `broken`.
`paused_at`	string	Conditional field returned when the automation is paused.
`version_number`	integer	Active automation version number. Pass it as `expected_version_number` when updating.
`last_ran_at`	string or null	Start time of the most recent run.
`govtribe_ai_project_name`	string or null	Name of the linked GovTribe AI project.
`created_at`	string	Automation creation timestamp.
`updated_at`	string	Automation update timestamp.
`total_runs_count`	integer	Total number of recorded runs for the automation.
`past_runs`	object array or null	Most recent run snapshots. Each item includes `id`, `state`, `is_user_initiated_test`, `started_at`, `finished_at`, and `memory`.
`broken_details`	object	Conditional field returned when the active automation version is broken. Includes `what_happened` and `how_to_fix`.

Start node fields

Every item in start_nodes describes one independent condition that can run the same saved prompt.

Field	Type	Notes
`start_node_id`	string	Stable ID used by `Update_Automation` start-node operations and multi-trigger `Test_Run_Automation` calls.
`trigger_type`	string	Trigger identifier such as `trigger-schedule` or `trigger-entity-activity`.
`trigger_config`	object	Public configuration accepted by create and update operations for this trigger type.
`summary`	string	Customer-facing description of the start condition.
Schedule details	mixed	Schedule items can include `frequency`, `timezone`, `rrule`, `time_of_day`, `day_of_week`, and `next_run_at`.
Saved-search details	mixed	Saved-search items include the scoped `saved_search`, independent lowercase `frequency`, and nullable `next_run_at`. The next run is present only while the workflow version is `published`. Their `trigger_config` includes `saved_search_model_id` and the same frequency.
Other event details	mixed	Other event items can include `target`, `pipeline`, `stage`, or `direction`, depending on the trigger type and scope.

For trigger-entity-activity, target contains the selected record's govtribe_type, govtribe_id, and display label. Its trigger_config contains the round-trippable target_govtribe_type and target_govtribe_id.

Follow-up Behavior

Search_Automations returns only govtribe_id when you omit fields_to_return.

When you request start_nodes, search also returns total_triggers_count. Search embeds up to three start nodes by default; raise max_triggers when the returned list is shorter than the total. Create and update responses return the complete ordered start-node list.

Use start_nodes[].start_node_id with start_node_changes when updating one trigger. Always send the current version_number as expected_version_number. Use the same stable ID as start_node_id when a manual test must choose among several start conditions.

trigger_details remains available for older integrations, but it describes only the first start node and is not a complete multi-trigger contract.

For trigger-saved-search-results, frequency is the automation trigger's polling cadence, not the saved search's email cadence. It remains unchanged when saved-search email preferences change. next_run_at is null while the workflow version is paused, suspended, broken, archived, or draft. Publishing schedules a fresh next run from the publication time; resuming also resets the listening checkpoint to the resume time, without replaying earlier results.

Request on_completion_notify_user and completion_notification_instructions before changing completion email behavior. Omit both fields on update to preserve the existing setting, or set on_completion_notify_user to never to turn completion emails off.

Request reasoning_level before changing it. Create defaults the field to low; update preserves the saved value when the field is omitted. A later message sent manually in the generated conversation uses the owner's personalization reasoning effort instead.

Search_Automations uses past_runs_limit to decide whether past_runs is present. Set past_runs_limit to 0 when you do not want embedded run history.

When an automation is paused or broken, GovTribe includes paused_at or broken_details even when those state-specific fields were not requested.

Example

{
  "govtribe_id": "AUTOMATION_ID",
  "govtribe_url": "https://govtribe.com/workspace/WORKSPACE_ID/oai-response/automations/AUTOMATION_ID",
  "name": "Award change and recompete monitor",
  "prompt": "Monitor this award for material changes and credible follow-on signals.",
  "reasoning_level": "low",
  "on_completion_notify_user": "conditionally",
  "completion_notification_instructions": "Email me only when the change affects capture timing or recommended action.",
  "start_nodes": [
    {
      "trigger_type": "trigger-entity-activity",
      "summary": "When \"Facilities support award (Award #W1234567890)\" changes",
      "target": {
        "govtribe_type": "federal_contract_award",
        "govtribe_id": "AWARD_ID",
        "label": "Facilities support award (Award #W1234567890)"
      },
      "start_node_id": "RECORD_CHANGE_START_NODE_ID",
      "trigger_config": {
        "target_govtribe_type": "federal_contract_award",
        "target_govtribe_id": "AWARD_ID"
      }
    },
    {
      "trigger_type": "trigger-schedule",
      "summary": "Mondays at 9:15 AM",
      "frequency": "weekly",
      "timezone": "America/New_York",
      "rrule": "RRULE:FREQ=WEEKLY;BYDAY=MO;BYHOUR=9;BYMINUTE=15",
      "time_of_day": "09:15",
      "day_of_week": "MO",
      "next_run_at": "2026-07-20T13:15:00Z",
      "start_node_id": "SCHEDULE_START_NODE_ID",
      "trigger_config": {
        "rrule": "RRULE:FREQ=WEEKLY;BYDAY=MO;BYHOUR=9;BYMINUTE=15",
        "rrule_timezone": "America/New_York"
      }
    },
    {
      "trigger_type": "trigger-saved-search-results",
      "summary": "When \"Healthcare IT recompetes\" saved search has new results (checked weekly)",
      "saved_search": {
        "govtribe_id": "SAVED_SEARCH_ID",
        "govtribe_type": "saved_search",
        "name": "Healthcare IT recompetes"
      },
      "frequency": "weekly",
      "next_run_at": "2026-07-23T14:00:00Z",
      "start_node_id": "SAVED_SEARCH_START_NODE_ID",
      "trigger_config": {
        "saved_search_model_id": "SAVED_SEARCH_ID",
        "frequency": "weekly"
      }
    }
  ],
  "total_triggers_count": 3,
  "trigger_details": {
    "trigger_type": "trigger-entity-activity",
    "summary": "When \"Facilities support award (Award #W1234567890)\" changes",
    "target": {
      "govtribe_type": "federal_contract_award",
      "govtribe_id": "AWARD_ID",
      "label": "Facilities support award (Award #W1234567890)"
    }
  },
  "state": "published",
  "version_number": 5,
  "last_ran_at": "2026-07-13T15:05:00Z",
  "govtribe_ai_project_name": "Recompete Watch",
  "created_at": "2026-07-01T13:15:00Z",
  "updated_at": "2026-07-13T14:00:00Z",
  "total_runs_count": 4
}

Response Shape

Start node fields

Follow-up Behavior

Example

On this page