> ## Documentation Index > Fetch the complete documentation index at: https://docs.intention.xyz/llms.txt > Use this file to discover all available pages before exploring further. # What's next > Capabilities being built after mainnet beta — what is committed, what is planned, and what each one unlocks. Two things are committed and in progress. Everything after them is planned work, grouped by what it serves. Status is marked as **committed** (being built now), **planned** (scoped, not started), or **exploring** (direction agreed, design open). ## Committed ### Trading agent runtime The execution loop an AI agent needs to trade on someone's behalf, rather than the loop a human interface assumes. A person can watch a screen and react. An agent cannot — it can only act on what the venue is able to prove. That gap is not closed by an API; it is closed by a runtime with defined semantics at every step: predict before committing, submit without ambiguity, resume after interruption, and read back a record that settles what happened. | Piece | What it provides | | ------------------------- | --------------------------------------------------------------------------------- | | **Pre-trade simulation** | Evaluate what an order would do against committed state before submitting it | | **Idempotent submission** | Retry an ambiguous submission without risking duplicate exposure | | **Named terminal states** | No outcome where the result is unknown — every action ends in a state with a name | | **Defined resume** | An agent that restarts knows where it was and what to do next | | **Attributed receipts** | What the agent did, under what authority, and what it cost — as protocol output | The account-model work sits alongside this: authorization today is all-or-nothing with an expiry, and richer terms — what an agent may hold, how much it may lose, what it may do next — are what turn a permission into a mandate. See [AI trading](/protocol/ai-trading). ### Institutional custody integration Direct funding from institutional custody, starting with Fireblocks. The blocker for institutional capital is rarely the venue's mechanics. It is that the capital cannot leave its custody arrangement to reach the venue. Integrating at the custody layer means an institution funds and trades without moving assets out of the controls it is required to operate under. ## Market surface | | Status | | -------------------------- | --------------------------------------------------------------------------------------------------------------------------------- | | **Spot markets** | Planned — the metadata layer carries spot definitions; the matching path is perpetual-only today | | **Broader RWA listings** | Planned — equities, indices, and commodities on the same execution model, with per-listing calendar and index-continuity handling | | **Portfolio margin** | Planned — margin computed across correlated positions rather than per market | | **Permissionless listing** | Exploring — a defined path for a market to be listed without a governance decision per asset | ## Liquidity and capital | | Status | | --------------------------------------- | --------------------------------------------------------------------------------------------------------------- | | **Vault market structure** | Planned — the capital structure around protocol-owned liquidity, including how third-party capital participates | | **Maker programs** | Planned — formal agreements and dedicated rates beyond the published rebate tiers | | **Cross-margin against RWA collateral** | Exploring | ## Access and accounts | | Status | | ------------------------------ | ---------------------------------------------------------------------------- | | **Richer authorization terms** | Committed — see the agent runtime above | | **Sub-account trees** | Planned — nested sub-accounts with independent books and risk parameters | | **Session and mobile flows** | Planned — scoped session credentials that do not require signing every order | | **Multi-signature accounts** | Exploring | ## Protocol | | Status | | ------------------------------- | ------------------------------------------------------------------------------- | | **Additional bridge routes** | Planned — chains and assets beyond the initial set | | **Validator set expansion** | Planned — growing the active set through mainnet | | **Fair-ordering constructions** | Exploring — reducing leader discretion over batch arrangement within a proposal | | **Governance surface** | Planned — the process by which risk parameters and listings are decided | ## Data and observability | | Status | | ------------------------------------- | -------------------------------------------------------------------------------------------------- | | **Historical data service** | Planned — reproducible archives with schemas, derived from committed state | | **Deterministic replay as a service** | Exploring — the venue as a training and backtesting environment, since every block replays exactly | Nothing here carries a date. Dates belong to [Milestones](/protocol/roadmap/milestones), which covers the two commitments that do have them. Items move from this page to the changelog when they ship. ## Where to go next What is running today, and the two dates that matter. Dated record of what has shipped. Why the agent runtime is the shape it is. What these capabilities are being built on.