Operator ruling wanted 2026-08-08: no self-hosted license exists and buying one is unwanted. Options fleet exploring: stay free-commercial-tier, switch/fork CE, wrapper-over-Plane, migrate to GitLab-native or OSS alternative, promote the markdown mirror to store-of-record. Decision memo lands in the MEMY-1121 campaign ledger (OPTIONS.md). Relates: MEMY-1121.

Operator rulings 2026-08-08

• Mirror-as-truth: GO, as terminus — reframed as a data-layer takeover. Truth = local .md cards feeding the global Frames postgres as a consistent Document variety (design card MEMY-1157); Plane demoted to auxiliary GUI/HCI over that data layer; the eventual primary UI is hand-rolled, not Plane's WUI and not a fork. Motivation is positive (use MORE tracker features, on our own schema), not just defect avoidance.
• Wrapper diff view + SID-minting fix: approved ASAP, not "this month". SID guard + audit trail already landed with Phase-0 (MR !157); wrapper M0+M2 = MEMY-1155; mirror git-init = MEMY-1156.
• OPTIONS decision points 3 (gate reconciler-deletion on verified-clean mirror) and 4 (keep Plane demoted): moot under the reframe — Plane stays up as the auxiliary GUI by construction; apply pragmatic-minimalism to any elaborate verification machinery rather than building gates for their own sake.
• Point 5: fork-CE / switch-CE / GitLab ruled out; Redmine ALSO ruled out (UI/featureset insufficient) — there is no third-party emergency-fallback web UI; the fallback is accelerating our own wrapper.
• myform: use encouraged — the plane dialect (Tiptap-conformant writer+reader) is carded as MEMY-1158, named by the operator as exactly myform's intended use case.
• Persistent heavy-dep server: the wrapper should be housed as our own single persistent process with room to load heavy deps (my-basis) — folded into MEMY-1155's housing requirements.