List Timesheet entries

The employee whose punches to read (employees.id, the Paylocity Employee ID). Required: Paylocity returns punch details per employee (its company-wide punch export is asynchronous and is not used). Exactly one employee per request: one employee id in employee[id], employee_id or employee, as a plain value, [eq], or a one-element list ([] or [in]); several employees or another form are refused with a 400 before Paylocity is called. Every entry's employee.id is this id. Paylocity notes that in a company set the Time & Labor employee number can differ from the HR & Payroll number: the punch record's own employeeId is kept in each entry's remote_data.shift.employeeId.

Filter by employee (employees.id).

Only these employees (employees.id, the HR WORKS person uuid). HR WORKS filters by username, so the ids are resolved through the person list Truto reads anyway; an id HR WORKS does not know returns an empty list without calling HR WORKS. Several values are supported: repeat the parameter, separate them with commas, use the [] form, or the operator forms [eq] / [in]. Any other form (for example [ne] or a nested object) is refused with a 400 instead of being ignored, so a filter Truto cannot express never returns unfiltered data. Truto resolves the person of every record by listing the company's persons once per page (GET /v2/persons, leavers included), because HR WORKS keys the response by a person identifier and the records themselves carry no person: the key pair therefore needs read permission for /persons as well. A key that no person matches, a key that two different persons match (one person's HR WORKS username can be another person's personnel number — HR WORKS keeps the two apart in nothing but their names) and a key pair without that permission all give the same answer: the record is returned WITHOUT employee, with the raw key in remote_data.truto_person_key. An employee is never guessed. HR WORKS only returns persons who have a time tracking regulation for the requested interval. HR WORKS renders stored working times as YYYYMMDD"T"HHMMSS"Z" but only documents UTC for the clock-in/out parameters, never for stored times, and has no time zone field anywhere. start_time, end_time and timezone are therefore left empty rather than guessed, and work_date is HR WORKS's own day bucket. hours_worked is HR WORKS's own figure for the day the entry sits in (workingTimeMinutes, which it reports next to a separate breakDurationMinutes) shared out over the entries of that day in proportion to their length — so a single entry simply gets the day's figure. Truto falls back to the difference between the entry's own two timestamps when HR WORKS reports no figure for the day, when an entry of the day is still open, or when the day also holds entries that are not working time (doctor's appointment, business errand and the like, which are not documented to be inside workingTimeMinutes). That fallback is only arithmetic on two strings whose zone HR WORKS does not state: a shift crossing a daylight-saving change is out by the change, and whether the difference already excludes an unpaid break is not documented either. The raw timestamps and the day's workingTimeMinutes / targetWorkingTimeMinutes / breakDurationMinutes are in remote_data.

Only this employee's entries (employees.id, Sesame employeeId); several employees in one call are refused (400). Accepts a plain value, a comma-separated list, repeated keys and [eq]/[in]; any other operator is refused (400) rather than dropped.

The employees whose timecards to read (employees.id). Required unless hyperfind is sent. UKG limits one request to 500 employees, 365 days and 28,000 employee-days; there is no paging, so split larger requests. Several UKG person IDs may be sent. Accepts one value, several comma-separated, a repeated key (x[]=a&x[]=b) or an eq/in filter; a filter form UKG cannot express is refused with 400 rather than ignored.

Filter by employee (employees.id).

Flat alternative to employee.id.

Flat alternative to employee.id (same values and forms).

Flat alternative to employee.id (same forms).

Flat alternative to employee.id.

Flat alternative to employee.id.

Flat alternative to employee.id.

approved returns approved punches, pending returns punches not yet approved (including open ones).

Filter by status: pending, approved, rejected (Deel declined) or processed (Deel processing and paid).

Filter by the status of the parent timesheet. Remote cannot filter timesheets by employee or date.

draft = Unsubmitted, pending = Submitted, approved = Approved, rejected = Denied or Sent Back. Repeat the parameter for several statuses. Workday has no processed status, so status=processed (or any other unsupported value) returns an empty list without calling Workday; unsupported values next to supported ones are ignored.

Required. The work dates (each entry's payroll date) to return, YYYY-MM-DD, both ends inclusive: work_date[gte] and work_date[lte] (gt/lt are treated as gte/lte), or work_date=YYYY-MM-DD (or work_date[eq]=YYYY-MM-DD) for one day. At most 29 days per request. Paylocity returns every shift that overlaps the time window it is asked for (an overnight shift is returned whole) and allows at most 31 days, and a shift can be booked to the payroll day before or after the one it was worked, so Paylocity is asked for one extra day on each side (relativeStart=<gte minus 1 day>T00:00:00, relativeEnd=<lte plus 1 day>T23:59:59, wall-clock times without a time zone as Paylocity requires) and only the entries whose work_date is inside work_date are returned: adjacent windows never return the same entry.

Required. The work dates to read (UKG applyDate, YYYY-MM-DD), as an object with gte and lte (gt/lt are treated as gte/lte), an eq, or a single date for one day. Both bounds are inclusive; UKG requires both. A value that is not a date, or a filter form this build cannot express, is refused with 400 rather than ignored.

Filter by business date (YYYY-MM-DD) with gte/lte. Takes priority over start_time.

Filter by work date (YYYY-MM-DD; Deel date_submitted, which Deel's create call defines as the date the work was performed). Accepts a date string (implies gte) or an object with gte/gt/lte/lt.

A date (YYYY-MM-DD), or an object with gte/gt, lte/lt or eq, sent as Sesame from/to (dates). gt and lt are treated as inclusive, eq as both bounds. Sesame does not document which date from/to compare (the clock-in date is assumed). Any other operator is refused (400) rather than dropped.

Filter by the date of the time block (YYYY-MM-DD). A plain date string returns that single day. Bounds are always inclusive (Workday fromDate/toDate): gt is treated as gte and lt as lte.

Start of the date interval to read (HR WORKS beginDate). Required. HR WORKS returns working times for a date interval and requires both ends. Truto always asks for interval=days, because the day bucket HR WORKS puts an entry in is the only work date it publishes; that shortens the maximum interval to 31 days, and a longer or negative one is refused with a 400 before calling. Accepts a date (YYYY-MM-DD), a date-time (only the date part is sent) or the operator forms [eq]/[gte]/[gt] and [eq]/[lte]/[lt]; gt and lt are treated as inclusive, because HR WORKS's window is inclusive. Several values, [ne] or a nested object are refused with a 400.

Filter by clock-in time. Accepts an ISO date-time (implies gte) or an object with gte/gt/lte/lt. Both bounds are inclusive: gt is sent as gte and lt as lte.

Return entries modified on or after this date-time. Accepts an ISO date-time string, or an object with gte or gt. The bound is inclusive: gt is treated as gte.

Entries updated in this window, sent as Sesame updatedAt[gte]/updatedAt[lte] (date-times with offset): a string (treated as gte) or an object with gte/gt, lte/lt or eq (gt and lt are treated as inclusive). A date alone covers the whole UTC day. Any other operator is refused (400).