Nobody talked you out — and they found the blind spot anyway

The frame HOLDS. No voice talked you out of the launch shape — a private company workspace (you + close friends, unpublished, unsold) plus an open contributor portal is normal, honest, and legally recognizable; work meetings others aren't invited to are exactly as reasonable as you said. Council #1's heavier armor was confirmed as over-engineered for what launches: keep the never-say list (it's also a labor-law hedge — see below), keep light visible summaries (what was decided, by whom — light, not meeting minutes), keep the open application path on the portal (cheap, builds trust). Defer the rest until team-hosting is actually sold.

deGrok's dissent, mined and kept sharp: "your 'normal company workspace' is not normal when the product is the social fabric itself… off-app gatherings among the close friends who also run the keeper business will be perceived as the real power center no matter how corporate your language." Not a reason to change the frame — a reason to do the three things below.

① German volunteer-vs-worker law — the hard one (5 voices, led by the European seats). Structured, recurring, directed testing work for a COMMERCIAL business in exchange for non-monetary "cool shit" can be classified as disguised employment (Mindestlohngesetz / Scheinselbstständigkeit territory) — and it gets sharper when paid employees and volunteers share one portal with different NDAs and upside. "A single bad-faith claim could crater the project." The fixes on the table: explicit beta-tester terms that disclaim employment (with the honest counter-voice: heavy legalism kills the playfulness — keep the terms light and human per your own copy law) · frame contributors as TESTERS, never task-directed workers · this is a COUNSEL question before the portal opens to strangers. ✅ SATISFIED AT LAUNCH (your round-8 word): the portal is NOT open to strangers — people you know personally only; your team stays ~5, family-only. Counsel still gates any future opening.

② The transition moment is the real risk window — not launch (Deepy & Seeky's sharpest point). When TeamPlex goes from "just Davi + friends" to sold hosting, the contributors who helped build it find out whether they're partners or labor. "Write the governance of that transition NOW, even if the code doesn't execute until later." A one-page transition covenant, written while it costs nothing.

③ Landlord optics on "the business is the keeper of the Social App location." The business owning the social commons reads as benevolent-dictatorship the moment anyone looks (the council split 50/50 on whether that ownership is protective or corrosive — both halves agreed it needs a stated commitment). Rudder's regulation-now: strict data segregation (alpha-tester data never mixes with contributor data) + a public commitment about what the business will never do with the commons.

Six as the team object: YES — it's already the house number (mush cap = 6, the Nook = 6 tiles; Kody1's geometry seat concurs). But two unanimous guardrails: ① ~5 teams per person = SOFT DEFAULT, never a hard cap — a hard cap breaks the person everyone wants, and (Deepy's point worth hearing whole) it breaks anyone whose teams are inherited, not chosen — family, clan, work, temple, school: "people overflow containers; the geometry is a container, not a cage." ② "Everybody has a team" must never be mandatory — the loner path is protected by your own ethos; solo practice stays a complete, first-class path, and "the geometry allows a single tile to occupy a zone alone — honor that." Rudder's shape: team object flexible 4–6, six preferred.

✅ Your round-8 addition — TEAMS OF SEVEN: "you got to be on your own team too" — 5–6 joined + your own = seven total, and seven tiles are the most that fit at a Penrose vertex (the Stretch). A fun coincidence, recorded. Five-teams stays a soft default, future — nothing builds yet.