Research prototype, not medical advice.

Arcanum session 1: decision log (D114–D125)

Compiled by the council clerk (A51) from the Decision sections of arcanum/session-1/topics/T1.md–T8.md (merged at c24b135) and from the charter's annex (merged at 566e350). One line per decision; the full wording is in the cited file. The old log (founding-council-log.md, D1–D113) is read-only history: an old decision binds only if session 1 re-issued it (D120, T7-A1-2).

Status values: decided · president (answered by the president).

IdDecisionStatusSource
D114Biological age is a vector (one axis per organ system and hallmark, each with a functional anchor); summary published, never an objective. A version passes only the six-test battery against sealed baselines; published scores labelled "retrospective, not blind"; a confirmed worsening blocks any 3×/progress claim and treatment request unless a sealed, tested simulation predicts a combination cancels it; test 4 "blind (key not isolated)" until the scoring environment exists.decidedarcanum/session-1/topics/T1.md
D115Simulation v0 = organ-level damage/repair model on the T1 axes; interventions mapped to parameters from mechanistic evidence only, sealed before lifespan is seen; split, strata and class labels in one seal record; scored against a sealed Gompertz + hazard-ratio null on three strata, each with its own sequential test; ranks no treatment before the metric clears the T1 battery.decidedarcanum/session-1/topics/T2.md
D116The president's arm: calibration and challenge of the twin, never validation of the metric; measurements only. First request = noise baseline (two draws + one blind split duplicate) with a sealed test-retest interval. Public repo gets forecast hashes, aggregates (≥ 5 resolved per measure), noise-ratio widths, one count of open twin issues; per-item records only for measures listed in consent.md. Storage location replaced by PR8.decidedarcanum/session-1/topics/T3.md
D117Cohort 1 = untreated UM-HET3, both sexes, testing sealed forecasts. Every mouse request is a CI-checked file (sealed prediction hash, refutation criterion, power script, dated novelty search, no open S1/S2). 3× (or 2×) claimed only after full survival, median and maximum vs in-range controls, replicated at a second site sealed before the first result. The 3× target never moves.decidedarcanum/session-1/topics/T4.md
D118Blind = sealed before any public trace of the forecast quantity; post-interim items name the interim curve, hash its extrapolated baseline, labelled "blind (post-interim)". A dated L1 trace search (≤ 7 days old) gates every seal. SLAM access = obstacle O16, owner A2, month 1.decidedarcanum/session-1/topics/T5.md
D119Fleet cap 9 active workers, ≤ 2 local builds (lock file), heavier runs in Actions; A10 re-measures monthly. Phase 1 and phase 2 worker splits as in T6. Verification = ceil(0.15 × cap); A7 and A11 start no worker without asking A1.decidedarcanum/session-1/topics/T6.md
D120Procedures: sealing P1–P5 and audits P1–P8 adopted as written (with the T7 amendments: main-only scoring environment, error definition, D51 per finding over rolling 3 months, unit = mission, exact binomial bounds for small pools). Mouse trust level = SYRCLE capped by T1-A2-3 flags. Re-issues the surviving old decisions listed in T7 (T7-A28-1/-2, T7-A10-1/-2, T7-A50-1/-4); anything not re-issued is dead.decidedarcanum/session-1/topics/T7.md
D121Obstacle register O1–O16 with owners, exits graded by Verification, month-≤ 6 milestones (T8-A28-4) and Platform dates (T8-A10-4).decidedarcanum/session-1/topics/T8.md
D122Session close form to the president: his data (location, agent read access, publication option), go/no-go on noise-baseline draws, repo settings only he can make, pending M2 re-seal. Answered 2026-10-20: see PR8 and PR9 in president-rulings.md.presidentarcanum/session-1/topics/T8.md; PR8, PR9
D123The charter is redrafted from D114–D122 on PR #28 and goes to D46 ratification.decidedarcanum/session-1/topics/T8.md
D124Re-issues from the old log: D18's refusals (no dosing or personal medical advice, no unpinned external data, no scoring of our own forecasts, no dual-use work); D43's trigger (separate agent identities re-proposed once more than 20 agents write); D77's Model: trailer (presence check only).decideddocs/charter/charter.md annex (566e350)
D125The charter changes only by a numbered decision voted in an Arcanum session; Verification diffs the amendment's citations before the president merges it; once D69 is live, CI also refuses a charter change that cites no decision on main.decideddocs/charter/charter.md annex (566e350)

D124 and D125 were voted by the four members in their D46 reviews of 1e69101 (D125 with A50's reservation, now part of its text) and confirmed by the president at the charter merge.