Our current house rules are:
* Fully vaccinated people are "invisible", at least in small groups -- they can come over and you can interact with them indoors freely.
* Gatherings where EVERY person is at least 2 weeks past the first dose, and it's small enough to verify this, are fine.

The rationale for rule #2 is: the numbers I've seen seem to roughly suggest that a fully vaccinated person interacting with an unvaccinated person has about the same transmission risk (~5%) as two people who are both two weeks past the first dose (20% * 20%).

Sticking with outdoor socialization for now, especially with people you don't know or can't verify the vaccine status of, seems like a pretty good idea. I expect we'll need to have more conversations about this once larger indoor gatherings seem like they might make sense.

Germany here. On my premise, I have a shared garden with three households, 3 kids in school (homeschooled every other week), 2 kids in kindergarden. Since we are very very lucky with this configuration (in terms of the kids being able to play with each other in the shared garden and not being stuck indoors all the time), we have had hard rules most of the time and everyone isolated on the premise, while most of the research on covid spreading was going on - this meant no kindergarden, no school, only one other household to meet adults with outdoors, FFP2-masked, with 2m distance.

While most risk factors on spreading are coming to a scientifically viable conclusion, the risks and length of long-term effects of a covid infection are not. Combined with the very low detection rate of covid in children before - which is bound to increase now through the mandated testing twice a week in most of germany - we are still being cautious (we also have people on the premise with preconditions, which are in the mid- to high-risk category for complications from an infection).

So our rules for now:

• Kids are allowed to meet other kids outside, that are in the same group they meet with anyway (same kindergarden group or school class)
• Meetings outside for adults are okay (due to the low risk of infection in the lower 0.x% range without mask, but with distance)
• Work-related meetings indoors that can't be virtual only under precautions (open windows, masked, distance)

The idea is to (still) restrict contacts to the same in-group and to minimise any contacts out-group. This works for small groups (our premise) as well as larger groups (premise and our contacts) as long as we hold the contacts accountable to the standard as well. This makes the risk more calculated than having no rules, but obviously more risky than our previous strict rules.

Since it's still unclear, if a two-shot vaccination (where the second shot is extremely important, since it boosts the antibody response by a factor of 20-40x) actually prevents a vaccinated person from being a spreader (with cautiously optimistic scientific finding towards "yes"), we will keep these rules for a little longer and treat vaccinated contacts similar to unvaccinated, until the findings are clearer. Which will hopefully be in the next weeks.

Thanks for writing this up! I'm also in the process of figuring out an agreement with roommates and was curious about this part as I expect we might not all have the same initial take: 

Shared spaces are a bit different. While we haven't decided for sure yet, I think it's probably going to end up being that you can invite anyone in who is fully vaccinated.

If I understand correctly, the tradeoff is between the benefits of everyone being to invite guests regardless of their vaccination status (and having one less 'rule' to think about) and the risk of 1) someone in the house contracting Covid 2) someone in the house transmitting Covid to people outside the house.

For your group, would you say it's primarily 1 or 2 that would push the agreement toward the version you describe? Are you considering similar agreements for other infectious diseases (I guess influenza is the most obvious one with a mix of vaccinated and unvaccinated people in the population)? Is your group considering any distinctions based on the reason someone hasn't been vaccinated?