Roadmap.

Date: 2026-10-13. Status: draft for founder review. Dates are targets, not commitments. Anything marked [FOUNDER GATE] cannot proceed without an explicit founder decision.

Owners are roles: Engine (indexer, scorer, merkle publisher, verifier), Launch (Meteora DBC adapter, treasury ops), Web (site and API), Brand (identity, video), Docs (this folder), Founder.

Phase summary

P0: Foundation (done)

Delivered: design spec, brand direction, this documentation set.

Exit criteria (met when the founder signs off on the docs and spec):

• SPEC.md approved under the founder's automation mandate.
• Documentation set drafted.
• Founder reviews the finished result (pending, see calendar week 1).

Owner: Docs, Founder. Risk: spec details (fee split, tenure curve) change after review; docs must be updated in the same change.

P1: Devnet proof (weeks 1 to 4)

Goal: prove the full loop on devnet with real transactions, no mock rows.

Work items:

1. Launch adapter: create a token on a Meteora Dynamic Bonding Curve with the shared fee config (Launch).
2. Indexer: ingest balances and swaps from Helius into Postgres (Engine).
3. Scorer: compute Dwell Scores per 24h epoch with the exact rules in TECHNICAL.md (Engine).
4. Merkle publisher: build tree, publish root and snapshot, fund the distributor (Engine, Launch).
5. Verifier CLI: independent recomputation from the public snapshot (Engine).
6. Web: token page, portfolio claim flow, leaderboard, public API v1 (Web).
7. Open design questions resolved and written into TECHNICAL.md (see "Open design questions" below).

Exit criteria:

• At least 3 consecutive devnet epochs scored, published and claimed by at least 5 distinct test wallets.
• Verifier CLI output matches the published root byte for byte for every epoch.
• Test suite covers: sell reset, buy no reset, transfer-in exclusion, creator exclusion, rounding remainder, rollover.
• Failure drills run: indexer restart mid epoch, RPC outage, wrong root detected.
• Open design questions closed.

Risks: payout in native SOL versus the distributor's SPL token model (see open questions); Helius rate limits; devnet liquidity not representative of mainnet.

[FOUNDER GATE] Founder reviews devnet proof before P2 starts.

P2: Closed mainnet pilot with caps (weeks 5 to 8)

Goal: real money, small and bounded, invited participants only.

Proposed caps (founder can tune):

• Max 5 pilot tokens, launched by invited creators only.
• Per token curve cap: graduation threshold stays at 85 SOL, no raise of it during pilot.
• Max pool funding per epoch per token: 5 SOL; anything above routes to the reserve until reviewed.
• Allowlisted wallets for launching; claim is open to any holder.
• Kill switch: Launch can pause new launches and epoch publication instantly.

Work items:

1. Complete every item in MAINNET-READINESS.md that is marked required for pilot.
2. Squads 2-of-3 vault created, signers confirmed, test transaction dry run.
3. Pilot creators onboarded with CREATORS.md.
4. Daily operations log and an on-call rota for the pilot weeks.

Exit criteria:

• At least 14 consecutive mainnet epochs published and verified without a restated root.
• 100 percent of pilot claims succeed or are explained in the incident log.
• No unresolved indexer dispute older than 72 hours.
• Security review findings of severity high or above closed.
• Legal review output received and addressed (see LEGAL-NOTES.md).

Risks: low volume means tiny payouts and weak signal; a single bug costs real funds; regulatory exposure begins when real holders are paid.

[FOUNDER GATE] Mainnet go decision before P2 starts. [FOUNDER GATE] Legal review sign-off before P2 starts. [FOUNDER GATE] Cap changes during P2.

P3: Public launch (weeks 9 to 10)

Goal: open launching to anyone, with limits that scale only after evidence.

Work items:

1. Open the launch page to all creators, with rate limits and abuse reporting.
2. Publish public docs, API docs and the verifier CLI.
3. Launch announcement and explainer video.
4. Monitoring and alerting on indexer lag, vault balance, root publication.

Exit criteria:

• Public launches run for 7 days with no Severity 1 incident.
• Indexer lag under 10 minutes at the 95th percentile.
• Support queue answered inside 24 hours.
• Launch-day load test passed at 5x pilot traffic.

Risks: spam tokens, impersonation tokens, scam creators, press attention before the system is stable; competitor copying.

[FOUNDER GATE] Public launch go decision. [FOUNDER GATE] Announcement copy approval (must contain no yield or return language).

P4: Growth (weeks 11 to 12 and beyond)

Candidate work, each needing a founder decision before commitment:

• Fee split tuning based on measured data.
• Creator tooling: analytics, scheduled announcements.
• Additional language support for docs.
• Integrations: more data feeds for terminals that can read the public API.
• Custom on-chain program (explicitly out of scope for v1; would need its own audit and budget).

Exit criteria for the first growth review:

• Retention report: share of scored wallets still holding after 14 days.
• Fee volume and payout totals reported from real data only.
• Decision memo on whether to continue, change, or stop each experiment.

Risks: chasing volume by loosening anti-sybil rules; adding features before the core is stable.

Open design questions (resolve in P1)

1. Native SOL versus wrapped SOL. The Merkle distributor handles SPL tokens. Payouts probably need wrapped SOL with an unwrap step in the claim UI. Confirm with the pinned distributor version.
2. Buy blending. SPEC says buys do not reset tenure. Because tenure is per wallet, a wallet could hold dust for 14 days then buy a large amount at full multiplier. Option: blend tenure by token weight on buys. Founder decision.
3. Leaf format. Must match the pinned distributor version exactly.

12-week calendar (from 2026-10-13)

Calendar risk: weeks 11 and 12 fall over end of year holidays. Plan for slower response and avoid risky changes then.