create-schedule

Schedule an agent or a process flow to run unattended, and return its schedule_id. Needs the fabric:orchestrate.write grant.

curl -sS -X POST https://fabric.wexa.ai/mcp/$FABRIC_PROJECT_ID \
  -H "authorization: Bearer $FABRIC_API_KEY" \
  -H "Mcp-Session-Id: $MCP_SESSION_ID" \
  -H "content-type: application/json" -H "accept: application/json" \
  -d '{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"create-schedule","arguments":{"agentflow_id":"6ab1945e7566f41b5c28e303","goal":"Created to capture delete-schedule, then deleted.","frequency":"0 9 * * 1"}}}'

{
  "content": [
    {
      "type": "text",
      "text": "{\"lifecycle_id\":\"qlc_000211\",\"result\":{\"schedule_id\":\"6ab196817566f41b5c28e341\",\"agentflow_id\":\"6ab1945e7566f41b5c28e303\",\"goal\":\"Created to capture delete-schedule, then deleted.\",\"frequency\":\"0 9 * * 1\",\"type\":\"recurring\",\"status\":\"active\",\"run_count\":0}}"
    }
  ],
  "isError": false
}

curl -sS https://fabric.wexa.ai/v1/create-schedule \
  -H "authorization: Bearer $FABRIC_API_KEY" \
  -H "content-type: application/json" \
  -d '{"agentflow_id":"6ab1945e7566f41b5c28e303","goal":"Created to capture delete-schedule, then deleted.","frequency":"0 9 * * 1"}'

{
  "lifecycle_id": "qlc_000211",
  "result": {
    "schedule_id": "6ab196817566f41b5c28e341",
    "agentflow_id": "6ab1945e7566f41b5c28e303",
    "goal": "Created to capture delete-schedule, then deleted.",
    "frequency": "0 9 * * 1",
    "type": "recurring",
    "status": "active",
    "run_count": 0
  }
}

import { Fabric } from '@wexa-fabric/sdk'

const fabric = new Fabric()
const result = await fabric.createSchedule({
  agentflow_id: '6ab1945e7566f41b5c28e303',
  goal: 'Created to capture delete-schedule, then deleted.',
  frequency: '0 9 * * 1'
})

// result is the payload itself, with no `.result` to read
{
  schedule_id: '6ab196817566f41b5c28e341',
  agentflow_id: '6ab1945e7566f41b5c28e303',
  goal: 'Created to capture delete-schedule, then deleted.',
  frequency: '0 9 * * 1',
  type: 'recurring',
  status: 'active',
  run_count: 0
}

from wexa import Fabric

fabric = Fabric()
out = fabric.create_schedule(
    agentflow_id="6ab1945e7566f41b5c28e303",
    goal="Created to capture delete-schedule, then deleted.",
    frequency="0 9 * * 1",
)

# out is the result, already unwrapped
{
    "schedule_id": "6ab196817566f41b5c28e341",
    "agentflow_id": "6ab1945e7566f41b5c28e303",
    "goal": "Created to capture delete-schedule, then deleted.",
    "frequency": "0 9 * * 1",
    "type": "recurring",
    "status": "active",
    "run_count": 0
}

Parameters

Parameter	Type	Required	Notes
`agentflow_id`	string	Required	Id of the agent or process flow to run, the same id `run-agent` and `run-process-flow` take. For an agent from `list-agents`, use the part of its id before the dot. A malformed id is refused. An agent or process flow in this project; a compound agent id is accepted whole (see identifiers).
`goal`	string	Required	What the agent or process flow should do on each run. Non-empty.
`end_at`	integer	Optional	Unix seconds, in the future, after which a recurring schedule stops firing. Recurring only.
`frequency`	string	Optional	Cron expression "min hour day month weekday", repeating. Supply exactly one of `frequency` or `run_at`.
`limit`	integer	Optional	Stop after this many runs. Recurring only.
`run_at`	integer	Optional	Unix seconds for a single run, in the future. Supply exactly one of `frequency` or `run_at`. Refused together with `limit`, `end_at` or `time_zone`.
`table`	object	Optional	Bind to a Data Table: each firing picks up the next matching row and substitutes it into the goal. Needs `table_id`; the filter shape is under More details.
`time_zone`	string	Optional	IANA zone the cron's wall-clock time was written in, e.g. Asia/Kolkata. Recurring only.

Show full response

A real response, captured from a live call to create-schedule:

{
  "lifecycle_id": "qlc_000211",
  "result": {
    "agentflow_id": "6ab1945e7566f41b5c28e303",
    "conclusion": null,
    "created_at": 1790023297.717658,
    "created_by": "6a79a770b6c128ccc24bfca7",
    "end_at": null,
    "frequency": "0 9 * * 1",
    "goal": "Created to capture delete-schedule, then deleted.",
    "last_execution_id": null,
    "limit": null,
    "run_at": null,
    "run_count": 0,
    "schedule_id": "6ab196817566f41b5c28e341",
    "status": "active",
    "time_zone": null,
    "type": "recurring",
    "updated_at": null
  }
}

// the object you get back IS the result — there is no `.result` to read
{
  agentflow_id: '6ab1945e7566f41b5c28e303',
  conclusion: null,
  created_at: 1790023297.717658,
  created_by: '6a79a770b6c128ccc24bfca7',
  end_at: null,
  frequency: '0 9 * * 1',
  goal: 'Created to capture delete-schedule, then deleted.',
  last_execution_id: null,
  limit: null,
  run_at: null,
  run_count: 0,
  schedule_id: '6ab196817566f41b5c28e341',
  status: 'active',
  time_zone: null,
  type: 'recurring',
  updated_at: null
}

// The TS SDK names these camelCase — lifecycleId, traceId — where Python uses
// lifecycle_id and trace_id. Both are non-enumerable: readable, but omitted by
// Object.keys() and JSON.stringify(). `result.lifecycleId` here is undefined.
result.lifecycleId // 'qlc_000211'

# a dict subclass; `result` is already unwrapped
{
 "agentflow_id": "6ab1945e7566f41b5c28e303",
 "conclusion": None,
 "created_at": 1790023297.717658,
 "created_by": "6a79a770b6c128ccc24bfca7",
 "end_at": None,
 "frequency": "0 9 * * 1",
 "goal": "Created to capture delete-schedule, then deleted.",
 "last_execution_id": None,
 "limit": None,
 "run_at": None,
 "run_count": 0,
 "schedule_id": "6ab196817566f41b5c28e341",
 "status": "active",
 "time_zone": None,
 "type": "recurring",
 "updated_at": None
}

out.lifecycle_id  # 'qlc_000211' — an attribute, not a key

More details

A schedule runs the agent or process flow with its goal each time it fires. The result is the schedule: its schedule_id, type (recurring or once), status (active), a run_count of 0 and the frequency you set. The schedule is attributed to the user whose key was used.

Each schedule counts its runs in run_count and records the latest in last_execution_id, which get-execution takes unchanged. Read it back with get-schedule, pause, resume or retire it with update-schedule, and remove it with delete-schedule.

Binding to a table. The table object takes a required table_id and an optional search_filters list. Each filter is {column, operator, value, type}: column is a column name or column_id as get-table reports it, operator is one of eq, gt, gte, in, lt, lte, ne, nin, value is an array for in and nin, and type is the column type, e.g. string, number or datetime.

A schedule is one way to start a run without a person asking. A trigger starts one on an event or a time set on an agent, connector or table.

Marked consequential: it is audited in full, and a policy rule may hold it for a person to approve. See Policy and approvals.

Every tool call works the same way: How every tool call works.