How to reconcile things that are measured at different rates is a tough one. I'm not an expert but I'll share here what I've seen in practice with other methods and how that relates to gimme. I welcome others to chime in here who might have more experience that I do at this.
I would suggest time-locking the observations by having HRV also be daily. More information is always better. If that is not doable or easy given your study design, then I would suggest one of these two options:
(1) You can mark it missing on the days where HRV is not taken. This means there will be HRV information every 7th day or so.
(2) You can input the HRV from a nearby HRV assessment. So, perhaps days 1-3 have the HRV from baseline (day 1), and days 4-10 use the HRV value obtained on day 7, and so on. Here, the "daily" HRV value is informed by the most recent value available.
Given those two options that I've seen in the literature, I would suggest #2 if you believe that a weekly assessment can inform what the HRV might be like on nearby days.