Roadmap

Regent moves through four phases: Preview, Testnet, Audit, Mainnet. Each phase has a goal, a list of deliverables and exit criteria. A phase ends when its exit criteria are met, not when a calendar says so.

Phase 1 Preview#

This is where the project is now.

Goal. Make the mechanism concrete enough to be criticised. A pitch cannot be checked. A simulator and a written specification can.

Deliverables.

• This site, with the design stated in full and its prior art credited. See Prior art and what is new.
• The preview: a simulator of the auction, the rent stream, the bell window and the duty monitor, running on simulated market data. No contracts are deployed, no funds move, and the only thing a wallet is ever asked for is a gasless typed-data signature marked as a preview.
• Tested TypeScript reference logic for the session clock, the auction, rent streaming and duties. Each module is written as the twin of a planned contract, and every number comes from one params file. See Parameters.
• Public docs, including draft contract interfaces in Architecture and contracts and an honest list of open questions in Edge cases and attack surface.

Exit criteria. The specification is stable enough to implement. The reference logic passes its tests, including daylight saving transitions, holidays and early closes. The open questions are written down, and each is assigned to the phase that must answer it.

Phase 2 Testnet#

Goal. Replace the simulator with real contracts and real bidders, without real money.

Deliverables.

• The seven planned contracts deployed on the Robinhood Chain testnet: SeatHook, SessionClock, SeatAuction, RentStream, BondRegistry, DutyMonitor and StakingVault.
• New test pools created with the hook, since hooks are fixed at pool creation.
• Real auctions with test funds, session after session, on the real calendar. Test tokens have no value.
• The site's on-chain data source switched on for testnet, reading contracts instead of the simulator.
• Parameter tuning from observed auctions: how bids behave around the anti-snipe window, how often seats go unsold, how the reserve moves.
• Every preview assumption resolved: the POST fee cap, the fail-safe default fee, the halt fee cap, when bidding opens, the scheduled close, the bond tiers, the slash rate and the oracle freshness threshold. Each becomes a specified value or a governed parameter.
• Answers to the open implementation questions: how fees reach the regent inside v4, how halts and sequencer downtime are detected, how the LP share of a slash is delivered.

Exit criteria. The contracts behave like the reference logic across all five session types, including a holiday, an early close and a daylight saving transition. No preview assumption is left. The code is frozen for audit.

Phase 3 Audit#

Goal. Have people who did not write the code try to break it.

Deliverables.

• One or more independent security audits of the frozen code. No auditor has been engaged yet, and none is named here until that is true.
• Fixes for the findings, and a re-review of the fixes.
• The audit report or reports published in full, including findings that were acknowledged and not fixed.
• A written description of any admin keys, upgrade paths and launch guards, so users can see who can change what.

Exit criteria. All critical and high severity findings are fixed and re-reviewed. The reports are public. If the fixes are large enough to change the design, the phase repeats.

There are no real deposits before this phase is complete. An audit reduces risk. It does not remove it. See Risks and disclaimers.

Phase 4 Mainnet#

Goal. Run real auctions for real LPs, starting small, and hand control to stakers.

Deliverables.

• A guarded launch: a small number of pools, with conservative limits that are lifted step by step. The specific guards are not designed yet and will be published before launch.
• New Regent pools on Robinhood Chain. Existing pools cannot adopt the hook, so these pools start empty and have to attract liquidity. The pitch is rent paid in advance.
• Monitoring that anyone can verify: auctions, rent streamed, breaches and slashes are all on-chain events.
• Governance handover to $RGNT stakers for the parameters the design makes governable: fee bounds, bell window length, tiers and the holiday calendar.

Exit criteria. Launch guards are lifted, the governable parameters are controlled by stakers, and any remaining admin powers are documented in public.

Mainnet does not change the honest summary for LPs: rent depends on auctions and can be zero, and deployed contracts carry risk of total loss.

What would change the plan#

The order of the phases will not change. Their content can. These are the things most likely to change it.

• Testnet auctions do not clear. If seats go unsold session after session, the auction format, the reserve rule or the session boundaries need rework before anything else happens.
• A v4 constraint. If the fee routing, the anti-JIT return delta or the rent donation cannot be built cleanly inside the hook, the architecture changes, and the docs change with it.
• Oracle or market-status coverage. Duties depend on a Chainlink tokenized-equity feed with market status for each stock. A pool without that coverage cannot have duties, and may not be worth launching.
• Chain changes. A change to how Robinhood Chain orders transactions would change the problem Regent solves, for better or worse.
• Audit findings. A design-level finding sends the project back to testnet.
• Legal and regulatory review. Stock Tokens are not available to US persons, and rules differ by jurisdiction. Legal review can change who can use what, or whether a phase happens at all.

Any of these can delay a phase or stop the project. That is the reason there are no dates.

Where to follow#

Progress is announced on X at x.com/Regent_rwa. This page is updated when a phase changes status. This site and that account are the only official channels.