Source of truth is log/entries.jsonl, append-only and validated on write. UPDATE-LOG.md and log/log.html are generated and must never be hand-edited — that is the failure mode this design exists to prevent. One command appends an entry and regenerates every view.
Question put by Chris: would a token project be easier to fund via private sale / presale? Easier to raise, yes — but declined on the specific facts. A retail-funded cap table is a permanent objection when selling trust-and-control infrastructure to finance teams, and a raise delivers capital without validation, which is the condition to avoid. Funding order of record: ecosystem grants (non-dilutive, legitimising) → revenue → equity if needed. Fundraising capability is held in reserve for Route 2, where a native settlement asset could be correct on the merits rather than as a funding workaround. Noted separately: the SEC's proposed Regulation Crypto Assets (under review Aug 2026) would allow a startup exemption up to $5M over four years with no accredited-investor requirement — worth re-checking if it is adopted, as it changes the compliant options materially.
Anonymous operation was normal in 2021-22 but is not viable in 2026 — it blocks institutional capital, ecosystem grants, business banking, and any licensed partner, since partner-licensed providers KYC principals. Aetherial's own published charter already refuses it: "it does not hide behind an anonymous team page." Settled by prior commitment.
Chris chose to run all three candidate models rather than pick one. They resolve into a single company: one verification engine sold to three buyers — Desk (advisory to launching teams), Signal (subscription intelligence for traders), Control (policy and verification for autonomous agent payments). They sequence by time-to-revenue, each funding the next: Desk earns in weeks with no build, Signal builds and hardens the engine paid for by users, Control sells that hardened engine to the buyer investors care about. Guardrails against the three-half-products failure: one shared engine with no per-surface pipelines, Desk capped at two live engagements, gate-based sequencing rather than parallel starts, and a single-product pitch to investors.