@serfab/cadre-core 1.13.0 → 1.14.0

package/src/membership-connection-gater.ts

@@ -1,570 +1,637 @@
-/**
- * Control-network inbound admission gates (defense-in-depth): the
- * encrypted-connection checkpoint and the circuit-relay reservation checkpoint.
- *
- * Layer 2 of the membership enforcement chain: the PRIMARY gates are per-stream
- * — every sensitive sereus control protocol rejects a peer that fails the
- * voucher-anchored membership check (wake and strand-addr via
- * `CadreNode.isAuthorizedMember`; the Optimystic control-DB protocols via the
- * materialized-snapshot gate `CadreNode.authorizeInboundControlStream`), and
- * the replicated rows an outsider *can* still write are disbelieved at read
- * time. This module adds the opportunistic connection-level layer on top: a
- * peer this node can positively determine is NOT authorized is refused at the
- * encrypted-connection checkpoint, before any protocol negotiation, so a
- * known-nothing outsider is never even in the conversation — EXCEPT on a node
- * that runs the circuit-relay server, where the refusal moves to the
- * relay-reservation checkpoint instead (see "The relay-reservation seam").
- *
- * ## The stranger allowlist (the ONE place it is defined)
- *
- * A libp2p connection gater decides per CONNECTION, before protocols are
- * negotiated, so "allow seed, deny repo" cannot be expressed here — instead the
- * policy admits a connection whenever a legitimate stranger interaction could be
- * riding it, and the per-stream gates take over. The complete set of protocols a
- * NOT-yet-authorized peer may legitimately speak on a control node is:
- *
- *  - `/sereus/seed/1.0.0` ({@link SEED_PROTOCOL}) — enrollment seed delivery.
- *    An owner dials a brand-new node to seed it (the new node has no members
- *    yet, so its gate is inert), and an invited phone dials the owner before it
- *    is authorized (admitted via the enrollment window `CadreNode.createInvite`
- *    opens). The handler's own trust decision is the anchored seed-trust policy.
- *  - `/sereus/formation/1.0.0` ({@link FORMATION_PROTOCOL}) — cross-party
- *    strand formation via open invitations. Stranger-facing BY DESIGN: the
- *    initiator is another party, and its token is only checkable inside the
- *    protocol. The exemption is therefore keyed on EXPECTATION of a stranger,
- *    not capability to serve one: stranger denial is suspended only while this
- *    node has at least one UNEXPIRED, NOT-FULLY-CONSUMED open invitation
- *    outstanding (`StrandSolicitationService.hasOutstandingInvitation` — the
- *    tokens this process minted or published, plus any still-redeemable
- *    `FormationInvite` row the usage recorder can see). Registering the
- *    responder does NOT suspend it, so every node — each registers one at
- *    `CadreNode.start` — keeps a live gate. The
- *    handler's own trust decision remains the per-token check, which is
- *    strictly finer than this one: a peer admitted here can still be rejected
- *    in-protocol for a bogus or spent token.
- *
- * There is one further connection-level carve-out, which is NOT a protocol
- * exemption and needs no stranger window:
- *
- *  - **Announced delegate peers** (`delegate-admission.ts`). A member's strand
- *    node runs as a separate libp2p identity whose peerId no sibling can
- *    recompute. Before starting a strand node, a member's control node
- *    announces that peerId over the already-authenticated strand-addr RPC, and
- *    the receiver holds a short-lived grant for it
- *    (`CadreNode.grantDelegateAdmission`). The grant admits the CONNECTION,
- *    admits the peer's RESERVATION at the seam below (without spending the
- *    unauthorized budget), and nothing else — it is deliberately invisible to
- *    the per-stream gate, so a delegate still gets refused on every
- *    members-only protocol.
- *
- * ## The relay-reservation seam
- *
- * A circuit-relay reservation is established by the reserving peer DIALING the
- * relay, so at the relay it is an inbound connection — and killing that
- * connection kills the reservation. For a genuine member whose `CadrePeer` row
- * has not yet replicated to this node, a connection-level deny here is NOT
- * self-healing the way a data connection's is: an outbound reconcile dial
- * re-establishes a data link, but no outbound dial can grant the REMOTE peer a
- * reservation, and a relay-only peer has no address of its own to dial back —
- * the reservation IS its address. Boot ordering makes the window ordinary (a
- * node's circuit listener runs inside `libp2p.start()`, before its own row
- * could have replicated anywhere), and a stalled replication makes it
- * unbounded. So the relay would be answering a RELAY question ("may this peer
- * use my forwarding capacity?") with a MEMBERSHIP answer, at a checkpoint where
- * a wrong answer is unrecoverable.
- *
- * On a node whose relay server is enabled the two questions are separated:
- *
- *  - The policy returns `'admit-for-relay'` instead of `'deny'`, and the
- *    connection is ADMITTED. The fail-closed per-stream gates still refuse such
- *    a peer every members-only protocol, so it gains identify/ping and the
- *    relay hop protocol, nothing else.
- *  - The reservation itself is decided at libp2p's
- *    `denyInboundRelayReservation` hook (the circuit-relay server consults it
- *    per RESERVE request): the policy admits members, delegates and configured
- *    infra outright, and admits peers it cannot place only within a bounded
- *    budget ({@link UnauthorizedReservationBudget}) — a member whose row is in
- *    flight always finds a slot under any sane cap, while outsiders cannot
- *    annex the party's relay capacity. COUNT and lifetime are the only bounds
- *    on such a peer: a party-run relay forwards without libp2p's per-connection
- *    data and duration limit by default (`relay-server.ts`), so what a granted
- *    slot carries is not capped.
- *  - An `'admit-for-relay'` connection that is NOT reserving is dropped: it has
- *    {@link RELAY_ADMISSION_RESERVE_DEADLINE_MS} to get a reservation ADMITTED
- *    at that hook, after which the gate CLOSES the underlying connection so
- *    both ends let go of the socket. The guarantee above thus weakens on
- *    relay-enabled nodes from "never in the conversation" to "in it briefly,
- *    and can speak nothing".
- *
- * The deadline clears on reservation ADMISSION, not on the reservation's own
- * success: this gate cannot observe the server-side `reserve()` outcome, so a
- * peer whose admitted reservation is then refused for server capacity keeps its
- * connection until either side closes it — bounded and mute, so harmless.
- *
- * Everything else a control node handles — the Optimystic control-DB protocols
- * (`/optimystic/control-<party>/{repo,cluster,sync,block-transfer}/…`), wake
- * (`/sereus/strand-wake/1.0.0`), and strand-addr (`/sereus/strand-addr/1.0.0`)
- * — is members-only, and every one of them enforces that per-stream. Wake and
- * strand-addr check `isAuthorizedMember` inside their handlers; the four
- * control-DB protocols are gated by the fail-closed
- * `CadreNode.authorizeInboundControlStream` (wired as libp2p's
- * `authorizeInboundStream` upstream hook), which judges each inbound stream
- * against the MATERIALIZED authorized-peer snapshot — synchronous and
- * in-memory, because a live DB read from inside the gate would deadlock into
- * mutual denial. The two layers complement, not duplicate: this connection
- * gate is fail-open over a live DB read (deny only on positive proof of an
- * outsider), the stream gate is fail-closed over the snapshot and has NO
- * stranger carve-outs — an enrollment window admits a stranger's connection
- * for seed delivery, yet its repo streams are still refused.
- *
- * ## The bring-up quiet period
- *
- * One state is decided BEFORE any of the above, and is not about the remote peer
- * at all: while this node's control-database bring-up is in flight
- * (`InboundAdmissionPolicy.bringUpInFlight`), the gate denies BOTH directions —
- * `denyDialPeer` and `denyInboundEncryptedConnection` — and then opens.
- *
- * The invariant it protects is "`ControlDatabase.initialize()` runs while this
- * node holds ZERO control connections". Building that database is a long chain
- * of cohort-consulting block probes, and every connected same-party sibling is in
- * the cohort those probes consult. A sibling that has not yet replicated this
- * node's `CadrePeer` row correctly refuses them at its own fail-closed per-stream
- * gate, so ONE connection in this window turns bring-up into
- * `BlockUnavailableError` and `start()` rejects. Retrying cannot converge: the
- * condition that would clear the refusal is this node's own row reaching the
- * sibling, and writing that row needs the database the retry is building. So the
- * only fix is to not be in the conversation yet — which is what this window is.
- *
- * The ordering is arranged so nothing SHOULD open a connection here anyway:
- * `network.relayAddrs` resolves to a listener that dials nothing and reserves
- * after bring-up (`relay-addrs.ts`), `controlNetwork.bootstrapNodes` is dialed
- * after bring-up (`CadreNode.dialControlBootstrapPeers`), and the control-cohort
- * reconcile pass is scheduled post-start. This window is what makes that a property
- * rather than an accident of ordering — the live cases it catches are peers the node
- * remembers from a previous run, which the connection manager auto-dials, and
- * inbound dials.
- *
- * A denial here costs a retry, not a partition: libp2p's connection manager
- * re-dials on its auto-dial cadence, the reservation supervisor re-drives, and a
- * denied inbound peer reconnects. The window opens on bring-up FAILURE too (via
- * `CadreNode.cleanup`), so teardown is never gated.
- *
- * ## Fail-open, deliberately
- *
- * Outside that window this layer only ever denies on a POSITIVE determination of
- * "unauthorized outsider while no stranger path is open". Any error, missing
- * dependency, or ambiguous state admits the connection (and the reservation) and
- * defers to the fail-closed stream gates — a DB hiccup must not partition a
- * legitimate cadre.
- */
-
-import debug from 'debug';
-import type { ConnectionGater, PeerId, MultiaddrConnection } from '@libp2p/interface';
-import { SEED_PROTOCOL } from './seed-bootstrap.js';
-import { FORMATION_PROTOCOL } from './strand-formation-protocol.js';
-import { withDeadline } from './control-stream.js';
-import { PARTY_RELAY_RESERVATION_TTL_MS } from './relay-server.js';
-import { ADMISSION_DECISION_TIMEOUT_MS, relayAdmissionReserveDeadlineMs } from './link-budget.js';
-
-const log = debug('sereus:cadre:connection-gater');
-
-/**
- * The protocols a not-yet-authorized peer may legitimately speak on a control
- * node (see the module doc above for why each is open and where its own trust
- * decision lives). Exported as the single reference point so the allowlist
- * cannot silently drift across modules.
- */
-export const STRANGER_OPEN_PROTOCOLS: readonly string[] = [SEED_PROTOCOL, FORMATION_PROTOCOL];
-
-/**
- * How long `CadreNode.createInvite` holds the inbound gate open for strangers
- * when the invite carries no explicit expiry: the invitee must dial in before
- * it is authorized, and an expiry-less invite gives the gate no bound of its
- * own. Redeeming after this window still works once the owner re-mints (or the
- * caller re-opens the window via `CadreNode.openEnrollmentWindow`).
- */
-export const DEFAULT_ENROLLMENT_WINDOW_MS = 30 * 60 * 1000;
-
-/**
- * How long an `'admit-for-relay'` connection may exist without a relay
- * reservation being ADMITTED at the `denyInboundRelayReservation` hook, after
- * which the gate closes the underlying connection. A reserving client asks for
- * its slot immediately after the connection upgrades (`relay-reservation.ts`
- * dials and requests in one drive; a strand node's configured circuit listener
- * does the same from inside its own `listen()`), so a connection idle past this
- * deadline is not reserving.
- *
- * Derived from the link ({@link relayAdmissionReserveDeadlineMs}): one link round trip for the
- * request to arrive plus one admission decision on it. This is the value at the default
- * declaration; `CadreNode` passes the one derived from the relay's OWN declaration, because the
- * relay is the machine that decides.
- */
-export const RELAY_ADMISSION_RESERVE_DEADLINE_MS = relayAdmissionReserveDeadlineMs();
-
-/**
- * Bound on the graceful close of an expired relay-only connection. It exists
- * because `AbstractMultiaddrConnection.close()` awaits an `idle`/`drain` event
- * when the connection still has unsent bytes, and that wait has no timeout of
- * its own — an unsignalled one never ends. This gate writes nothing to a
- * stranger, so the wait is not reachable today; the bound is here so a timer
- * callback can never hold an unending await.
- *
- * Applied through {@link withDeadline}, not `AbortSignal.timeout`: the latter is
- * not reliably present on React Native/Hermes, which loads this same module.
- */
-// eslint-disable-next-line no-restricted-syntax -- link-independent: bounds a local connection close inside a timer callback, and expiry aborts the connection instead
-export const RELAY_ADMISSION_CLOSE_TIMEOUT_MS = 2_000;
-
-/**
- * Default cap on concurrent relay reservations held by peers the membership
- * check could not place (see {@link UnauthorizedReservationBudget}). Small on
- * purpose: it exists for the handful of genuine members whose rows are still
- * in flight, not as public relay capacity. Overridable per node via
- * `network.unauthorizedRelayReservationCap` (0 refuses every unauthorized
- * reservation — the strict pre-seam posture).
- *
- * NOTE: this cap shares the relay server's own reservation store, whose
- * party-run default size is `PARTY_RELAY_MAX_RESERVATIONS` (128,
- * `relay-server.ts`); unplaced peers may therefore occupy up to this many of
- * those slots, and the rest are what members and delegates compete for. The two
- * are kept apart by hand — if this is raised, or an embedder shrinks the store
- * through `network.relayServerInit.reservations.maxReservations`, keep this one
- * well under the store's, or a fleet of unplaceable peers can crowd genuine
- * members out of the server's own store (which the gate cannot override).
- */
-export const MAX_UNAUTHORIZED_RELAY_RESERVATIONS = 8;
-
-/**
- * The connection-level outcome of the admission policy:
- *  - `'admit'` — a peer with a legitimate claim on the connection (member,
- *    delegate, infra, open stranger window, or any fail-open state).
- *  - `'deny'` — positively an outsider, and this node runs no relay server, so
- *    the connection can carry nothing legitimate.
- *  - `'admit-for-relay'` — positively an outsider (or an unreplicated member —
- *    indistinguishable), but this node runs the relay server: admit the
- *    connection so a reservation can be asked for, decide at the reservation
- *    seam, and drop the connection if no reservation is admitted in time.
- */
-export type InboundConnectionVerdict = 'admit' | 'deny' | 'admit-for-relay';
-
-/**
- * The admission decisions the gate defers to — implemented by
- * `CadreNode.admitInboundControlConnection` /
- * `CadreNode.admitControlRelayReservation` (enrollment windows, anchor state,
- * bootstrap infra, delegate grants, the authorized-member set, the
- * unauthorized-reservation budget). Kept injectable so the gater's
- * composition/fail-open behavior is unit-testable without a full node.
- */
-export interface InboundAdmissionPolicy {
-  /** Verdict on an inbound encrypted connection from this peer. */
-  admitInbound(remotePeerId: string): Promise<InboundConnectionVerdict> | InboundConnectionVerdict;
-  /** Should this peer be granted a circuit-relay reservation slot? */
-  admitRelayReservation(remotePeerId: string): Promise<boolean> | boolean;
-  /**
-   * True while this node's control-database bring-up is in flight — the
-   * {@link createMembershipConnectionGater} quiet period (see the module doc).
-   * Peer-independent, unlike the two decisions above, and SYNCHRONOUS: it is
-   * read on the outbound-dial path, where an await would be a new stall.
-   *
-   * Optional — a policy that omits it is never quiet, which is the right default
-   * for every caller that is not a booting `CadreNode`.
-   */
-  bringUpInFlight?(): boolean;
-}
-
-/**
- * Bounded budget of concurrent relay reservations for peers the membership
- * check could not place. `tryAdmit` is the whole surface: an already-admitted
- * peer refreshes its entry (a reservation refresh never double-counts), a new
- * peer takes a free slot or is refused. Entries expire after `ttlMs`, which
- * `CadreNode` passes as the relay server's own resolved `reservationTtl`
- * (`relay-server.ts`): the server holds an unrefreshed reservation exactly that
- * long, and a live reserver re-requests (re-hitting the admission hook,
- * refreshing its entry) well before expiry — so the live entry count tracks the
- * server's own occupancy without this module reaching into the server's
- * reservation store. Injectable `now` keeps expiry testable without fake
- * timers. Authorized members and delegates are never run through this — the
- * policy admits them before consulting the budget, and `release`s the slot such
- * a peer took while it was still unplaceable, so the boot-ordering window a
- * member passes through costs the budget nothing once its row lands.
- */
-export class UnauthorizedReservationBudget {
-  private readonly admitted = new Map<string, number>();
-
-  constructor(
-    /** Most peers admitted at once — `network.unauthorizedRelayReservationCap`, resolved. */
-    readonly cap: number = MAX_UNAUTHORIZED_RELAY_RESERVATIONS,
-    private readonly ttlMs: number = PARTY_RELAY_RESERVATION_TTL_MS
-  ) {}
-
-  /** Number of live (unexpired at last prune) entries — test/diagnostic surface. */
-  get size(): number {
-    return this.admitted.size;
-  }
-
-  /** Admit (or refresh) `remotePeerId` if a slot is free; false when the cap is spent. */
-  tryAdmit(remotePeerId: string, now: number = Date.now()): boolean {
-    this.prune(now);
-    if (!this.admitted.has(remotePeerId) && this.admitted.size >= this.cap) {
-      log('Unauthorized-reservation budget spent (%d/%d) — refusing %s', this.admitted.size, this.cap, remotePeerId);
-      return false;
-    }
-    this.admitted.set(remotePeerId, now + this.ttlMs);
-    return true;
-  }
-
-  /**
-   * Give back the slot `remotePeerId` holds, if any — called when the peer has
-   * become admissible on its own merits (its membership row replicated, or a
-   * delegate grant landed), so the slot it took during the boot-ordering window
-   * does not stay spent for the remaining TTL on a peer that no longer needs it.
-   */
-  release(remotePeerId: string): void {
-    this.admitted.delete(remotePeerId);
-  }
-
-  private prune(now: number): void {
-    for (const [peerId, expiresAt] of this.admitted) {
-      if (expiresAt <= now) {
-        this.admitted.delete(peerId);
-      }
-    }
-  }
-}
-
-/**
- * Build the control node's connection gater: the caller-supplied gater (if any)
- * with membership admission composed onto `denyInboundEncryptedConnection` —
- * the earliest checkpoint where the remote's authenticated PeerId is known —
- * and reservation admission composed onto `denyInboundRelayReservation` — the
- * hook the circuit-relay server consults per RESERVE request (inert on a node
- * whose relay server is off; libp2p never calls it there).
- *
- * `denyDialPeer` is composed too, but ONLY for the bring-up quiet period (see
- * the module doc): outside that window outbound dials are never gated by
- * membership — this node decides for itself who to talk to.
- *
- * Composition semantics: every hook of `base` is preserved as-is; on the three
- * composed hooks a deny from EITHER the base gater or the admission policy
- * denies. A policy error — or a decision slower than `decisionTimeoutMs` —
- * takes the fail-open outcome (connection admitted / reservation admitted, see
- * module doc); the base gater's verdict is still honored first.
- *
- * An `'admit-for-relay'` verdict admits the connection and arms a
- * `reserveDeadlineMs` timer against it; the timer is disarmed when a
- * reservation for that peer is ADMITTED at the reservation hook, and closes the
- * underlying `MultiaddrConnection` when it fires first.
- *
- * NOTE: `base` is spread, so a gater passed as a CLASS INSTANCE would lose its
- * prototype methods; every caller in this repo (and libp2p's own default)
- * supplies a plain object. If a class-based gater ever shows up, delegate
- * per-hook instead of spreading.
- *
- * Deny timing, as observed by the denied dialer: noise negotiates the muxer in
- * the security handshake's early data, so the DIALER's upgrade may complete
- * (its `dial()` resolves) before this receiver-side hook runs. The deny then
- * aborts the receiver's upgrade — the receiver never registers the connection
- * and never creates its muxer, so no protocol can ever be negotiated — and the
- * dialer sees its "open" connection close moments later.
- *
- * Control node only: strand cohort nodes legitimately connect cross-party
- * peers, so cadre membership never gates them. An OPEN strand's node receives
- * the raw configured gater unchanged; a CLOSED strand's node composes its own
- * revoked-peer denial onto it instead — a deny-list on positive revocation
- * evidence, not this module's membership admission — see
- * `strand-revocation-enforcer.ts`.
- */
-export function createMembershipConnectionGater(
-  policy: InboundAdmissionPolicy,
-  base?: ConnectionGater,
-  decisionTimeoutMs: number = ADMISSION_DECISION_TIMEOUT_MS,
-  reserveDeadlineMs: number = RELAY_ADMISSION_RESERVE_DEADLINE_MS
-): ConnectionGater {
-  const pendingReservations = new PendingReserveDeadlines(reserveDeadlineMs);
-  const quiet = (): boolean => policy.bringUpInFlight?.() ?? false;
-  return {
-    ...base,
-    denyDialPeer: async (peerId: PeerId): Promise<boolean> => {
-      if (await base?.denyDialPeer?.(peerId)) {
-        return true;
-      }
-      if (quiet()) {
-        log('Control-database bring-up in flight — refusing the outbound dial to %s', peerId.toString());
-        return true;
-      }
-      return false;
-    },
-    denyInboundEncryptedConnection: async (peerId: PeerId, maConn: MultiaddrConnection): Promise<boolean> => {
-      if (await base?.denyInboundEncryptedConnection?.(peerId, maConn)) {
-        return true;
-      }
-      if (quiet()) {
-        log('Control-database bring-up in flight — refusing the inbound connection from %s', peerId.toString());
-        return true;
-      }
-      const remotePeerId = peerId.toString();
-      let verdict: InboundConnectionVerdict;
-      try {
-        verdict = await decideWithinDeadline(
-          () => policy.admitInbound(remotePeerId), 'admit', decisionTimeoutMs, `admitInbound(${remotePeerId})`
-        );
-      } catch (error) {
-        log('admitInbound threw for %s — admitting (fail-open; stream gates decide): %o', remotePeerId, error);
-        return false;
-      }
-      if (verdict === 'admit-for-relay') {
-        pendingReservations.arm(remotePeerId, maConn);
-        return false;
-      }
-      return verdict === 'deny';
-    },
-    denyInboundRelayReservation: async (peerId: PeerId): Promise<boolean> => {
-      if (await base?.denyInboundRelayReservation?.(peerId)) {
-        return true;
-      }
-      const remotePeerId = peerId.toString();
-      let admitted: boolean;
-      try {
-        admitted = await decideWithinDeadline(
-          () => policy.admitRelayReservation(remotePeerId), true, decisionTimeoutMs, `admitRelayReservation(${remotePeerId})`
-        );
-      } catch (error) {
-        log('admitRelayReservation threw for %s — admitting (fail-open): %o', remotePeerId, error);
-        admitted = true;
-      }
-      if (admitted) {
-        pendingReservations.disarm(remotePeerId);
-      }
-      return !admitted;
-    }
-  };
-}
-
-/**
- * Run one admission decision under `timeoutMs`, resolving to `fallback` (the
- * fail-open outcome) if it has not settled in time — see
- * {@link ADMISSION_DECISION_TIMEOUT_MS} for why an unbounded await is not safe
- * on the connection hook. The timer is always cleared so a decided call never
- * holds the event loop open. Exported for the strand revoked-peer gate
- * (`strand-revocation-enforcer.ts`), which bounds its connection hooks with the
- * same fail-open contract.
- */
-export async function decideWithinDeadline<T>(
-  decide: () => Promise<T> | T,
-  fallback: T,
-  timeoutMs: number,
-  what: string
-): Promise<T> {
-  let timer: ReturnType<typeof setTimeout> | undefined;
-  try {
-    return await Promise.race([
-      Promise.resolve().then(decide),
-      new Promise<T>((resolve) => {
-        timer = setTimeout(() => {
-          log('%s exceeded %dms — taking the fail-open outcome (stream gates decide)', what, timeoutMs);
-          resolve(fallback);
-        }, timeoutMs);
-      })
-    ]);
-  } finally {
-    clearTimeout(timer);
-  }
-}
-
-/** One armed not-reserving deadline: the connection it will close, and its timer. */
-interface PendingReserveDeadline {
-  maConn: MultiaddrConnection;
-  timer: ReturnType<typeof setTimeout>;
-}
-
-/**
- * The not-reserving deadlines for `'admit-for-relay'` connections, keyed by
- * remote peerId (a Set per peer — one peer can hold several in-flight
- * connections). `arm` starts a timer that closes the connection; `disarm`
- * (called when a reservation for that peer is admitted) cancels every pending
- * timer for the peer. Timers are unref'd so an armed deadline never holds a
- * process open, and a timer that fires against an already-closed connection is
- * a no-op (`close()` returns at once on any status but `open`).
- *
- * Disarming is final for the connections it cancelled: a peer that reserved
- * once and then lets its reservation lapse keeps a mute connection (the
- * fail-closed stream gates refuse it every members-only protocol) until either
- * side closes it. Bounded — only a peer the reservation policy already admitted
- * can reach that state, so unplaceable peers are bounded by the budget cap.
- */
-class PendingReserveDeadlines {
-  private readonly byPeer = new Map<string, Set<PendingReserveDeadline>>();
-
-  constructor(private readonly deadlineMs: number) {}
-
-  arm(remotePeerId: string, maConn: MultiaddrConnection): void {
-    const entry: PendingReserveDeadline = {
-      maConn,
-      timer: setTimeout(() => this.expire(remotePeerId, entry), this.deadlineMs)
-    };
-    (entry.timer as { unref?: () => void }).unref?.();
-    let entries = this.byPeer.get(remotePeerId);
-    if (!entries) {
-      entries = new Set();
-      this.byPeer.set(remotePeerId, entries);
-    }
-    entries.add(entry);
-    log('Admitted %s for relay only — dropping the connection unless a reservation is admitted within %dms', remotePeerId, this.deadlineMs);
-  }
-
-  disarm(remotePeerId: string): void {
-    const entries = this.byPeer.get(remotePeerId);
-    if (!entries) {
-      return;
-    }
-    this.byPeer.delete(remotePeerId);
-    for (const entry of entries) {
-      clearTimeout(entry.timer);
-    }
-  }
-
-  private expire(remotePeerId: string, entry: PendingReserveDeadline): void {
-    const entries = this.byPeer.get(remotePeerId);
-    entries?.delete(entry);
-    if (entries?.size === 0) {
-      this.byPeer.delete(remotePeerId);
-    }
-    log('Relay-only admission expired for %s — no reservation admitted within %dms, closing the connection', remotePeerId, this.deadlineMs);
-    void this.drop(remotePeerId, entry.maConn);
-  }
-
-  /**
-   * End an expired relay-only connection: a bounded graceful close, with
-   * `abort()` only as the escalation. The order cannot be reversed — `abort()`
-   * marks the connection `aborted`, and `close()` returns immediately on any
-   * status that is not `open`, so aborting first would make the close a silent
-   * no-op. Closing a connection the peer already dropped returns at once
-   * without throwing, so the fallback stays unreached in the ordinary case.
-   */
-  private async drop(remotePeerId: string, maConn: MultiaddrConnection): Promise<void> {
-    try {
-      await withDeadline(
-        RELAY_ADMISSION_CLOSE_TIMEOUT_MS,
-        `relay-only close for ${remotePeerId}`,
-        (signal) => maConn.close({ signal })
-      );
-      return;
-    } catch (error) {
-      log('Closing the expired relay-only connection from %s failed — aborting: %o', remotePeerId, error);
-    }
-    // NOTE: this fallback cannot actually free a WebSocket. `@libp2p/websockets`'
-    // `sendReset()` calls `websocket.close(1006)`, and 1006 is a reserved code
-    // RFC 6455 forbids an endpoint from sending, so `ws` throws and
-    // `AbstractMessageStream.abort()` swallows it — the local end flips to
-    // `aborted` while the socket stays up. Measured against
-    // @libp2p/websockets@10.1.3 (see tickets/blocked/report-libp2p-websockets-abort-close-code.md).
-    // `MultiaddrConnection` exposes no way down to the raw socket, so this is as
-    // far as this layer can go. Revisit — drop the fallback, or go back to a
-    // plain `abort()` — once that transport sends a legal reset code.
-    try {
-      maConn.abort(new Error(`relay-only admission expired: no relay reservation admitted within ${this.deadlineMs}ms`));
-    } catch (error) {
-      log('Aborting the expired relay-only connection from %s threw: %o', remotePeerId, error);
-    }
-  }
-}
+/**
+ * Control-network inbound admission gates: the encrypted-connection checkpoint
+ * and the circuit-relay reservation checkpoint.
+ *
+ * The fail-closed layer of the membership enforcement chain is per-stream: the
+ * control node's protocol guard (`control-protocol-guard.ts`) refuses a peer that
+ * is not an authorized member on every protocol not declared open to strangers,
+ * before its handler runs, and the replicated rows an outsider *can* still write
+ * are disbelieved at read time. This module is NOT a stranger boundary on an
+ * enrolled node — outside the bring-up quiet period it admits every inbound
+ * connection. Its jobs are:
+ *
+ *  - the bring-up quiet period (below), the one state in which it denies;
+ *  - deciding which admitted peers carry a deadline: a peer the policy cannot
+ *    place is admitted PROVISIONALLY, and its connection is closed at the
+ *    deadline unless the policy admits it by then (see "Provisional admission");
+ *  - handing circuit-relay reservations to the reservation policy (see "The
+ *    relay-reservation seam").
+ *
+ * Why a stranger's connection is admitted at all: the gate sees only a peer id,
+ * after the encrypted handshake and before any protocol, so it cannot tell an
+ * outsider from a device carrying a cadre invitation this node has not received
+ * yet, or from a sibling whose membership row is still replicating. Refusing
+ * would fail those at the connection; admitting costs a mute connection for a
+ * bounded time, because the protocol guard refuses it every members-only protocol.
+ *
+ * ## Admitted outright
+ *
+ * A peer the policy CAN place gets no deadline. Besides members, configured
+ * infrastructure and an un-enrolled node's whole world (empty anchor or empty
+ * authorized set), that covers the stranger windows: states in which this node
+ * expects a stranger on one of the `stranger-open` protocols declared in
+ * `controlProtocolClasses` (`control-protocol-guard.ts`):
+ *
+ *  - `/sereus/seed/1.0.0` — enrollment seed delivery.
+ *    An owner dials a brand-new node to seed it (the new node has no members
+ *    yet, so it admits everyone outright). The handler's own trust decision is the
+ *    anchored seed-trust policy — or, on a node started with a claim secret
+ *    (`CadreNodeConfig.claim`), the claim-secret policy. Such a node is the one
+ *    stranger-facing state that is NOT "admit the connection and let the
+ *    per-stream gates sort it out": while it waits to be claimed
+ *    (`CadreNode.isAwaitingClaim`) its connection is admitted, because the claim
+ *    seed has to ride one, but its control-DB streams and relay reservations are
+ *    refused ahead of every admission below, the empty-anchor one included. An
+ *    unclaimed node has no siblings to replicate from, and one on a public
+ *    address must not relay for anyone. The claim anchors the owner, after which
+ *    the ordinary rules apply.
+ *  - `/sereus/formation/1.0.0` — cross-party
+ *    strand formation via open invitations. Stranger-facing BY DESIGN: the
+ *    initiator is another party, and its token is only checkable inside the
+ *    protocol. The window is keyed on EXPECTATION of a stranger, not capability
+ *    to serve one: it is open only while this node has at least one UNEXPIRED,
+ *    NOT-FULLY-CONSUMED open invitation outstanding
+ *    (`StrandSolicitationService.hasOutstandingInvitation` — the tokens this
+ *    process minted or published, plus any still-redeemable `FormationInvite`
+ *    row the usage recorder can see). Registering the responder does NOT open
+ *    it, so on every node — each registers one at `CadreNode.start` — a stranger
+ *    with no invitation in play stays on the provisional deadline. The
+ *    handler's own trust decision remains the per-token check, which is
+ *    strictly finer than this one: a peer admitted here can still be rejected
+ *    in-protocol for a bogus or spent token.
+ *  - `/sereus/cadre-invite/1.0.0` — redemption
+ *    of a cadre invitation at any member machine. Stranger-facing BY DESIGN:
+ *    the device is not a member until the redemption writes its row, and its
+ *    proof of possession (a signature with the invitation's private key) is
+ *    only checkable inside the protocol. Keyed on expectation, as formation's
+ *    is: open only while this node holds at least one LIVE `CadreInvite` row
+ *    (`ControlDatabase.hasLiveCadreInvite`: not withdrawn, unexpired, uses left,
+ *    issuer still an owner). A member that has not yet received the device's
+ *    row by replication admits the device provisionally instead, seats the row
+ *    from the invitation bundle during the redemption, and so admits it at the
+ *    deadline's re-check.
+ *
+ * There is one further connection-level carve-out, which is NOT a protocol
+ * window:
+ *
+ *  - **Announced delegate peers** (`delegate-admission.ts`). A member's strand
+ *    node runs as a separate libp2p identity whose peerId no sibling can
+ *    recompute. Before starting a strand node, a member's control node
+ *    announces that peerId over the already-authenticated strand-addr RPC, and
+ *    the receiver holds a short-lived grant for it
+ *    (`CadreNode.grantDelegateAdmission`). The grant admits the CONNECTION
+ *    outright, admits the peer's RESERVATION at the seam below (without
+ *    spending the unauthorized budget), and nothing else — it is deliberately
+ *    invisible to the protocol guard, so a delegate still gets refused on every
+ *    members-only protocol.
+ *
+ * ## Provisional admission
+ *
+ * A peer the policy cannot place (`'admit-provisionally'`) gets its connection
+ * and a deadline — {@link PROVISIONAL_ADMISSION_DEADLINE_MS} at the default
+ * declaration; `CadreNode` passes the one derived from its OWN declared link,
+ * because the member is the machine that decides. At the deadline the gate asks
+ * the policy again:
+ *
+ *  - `'admit'` — the peer has since become a member, holds a delegate grant, or
+ *    this node now holds a live invitation (a redeeming device's row was
+ *    seated): the connection stays, and nothing is re-armed;
+ *  - `'admit-provisionally'` — the connection is closed (a bounded graceful
+ *    close, with an abort only as the escalation);
+ *  - the question throws or outlasts the decision deadline — the connection
+ *    stays (see "Fail-open, deliberately").
+ *
+ * A connection that is already gone by then costs no question. One peer may hold
+ * several provisional connections; each deadline re-asks on its own.
+ *
+ * NOTE: the gate does not stop a provisionally admitted stranger from being
+ * counted into this node's FRET ring and control-database cohort by opening or
+ * advertising the party's protocols — FRET and Optimystic decide that from the
+ * protocols libp2p records the peer as serving. Ticket
+ * `stranger-joins-the-control-cohort-by-advertising-protocols` (blocked).
+ *
+ * ## The relay-reservation seam
+ *
+ * A circuit-relay reservation is established by the reserving peer DIALING the
+ * relay, so at the relay it rides an inbound connection — and closing that
+ * connection kills the reservation. For a genuine member whose `CadrePeer` row
+ * has not yet replicated to this node that would NOT be self-healing the way a
+ * data connection's close is: an outbound reconcile dial re-establishes a data
+ * link, but no outbound dial can grant the REMOTE peer a reservation, and a
+ * relay-only peer has no address of its own to dial back — the reservation IS
+ * its address. Boot ordering makes the window ordinary (a node reserves at the
+ * end of its own `start()`, before its row could have replicated anywhere), and
+ * a stalled replication makes it unbounded. So the reservation is decided on its
+ * own terms, not by the connection's:
+ *
+ *  - It is decided at libp2p's `denyInboundRelayReservation` hook (the
+ *    circuit-relay server consults it per RESERVE request): the policy admits
+ *    members, delegates and configured infra outright, and admits peers it
+ *    cannot place only within a bounded budget
+ *    ({@link UnauthorizedReservationBudget}) — a member whose row is in flight
+ *    always finds a slot under any sane cap, while outsiders cannot annex the
+ *    party's relay capacity. COUNT and lifetime are the only bounds on such a
+ *    peer: a party-run relay forwards without libp2p's per-connection data and
+ *    duration limit by default (`relay-server.ts`), so what a granted slot
+ *    carries is not capped.
+ *  - An ADMITTED reservation disarms every provisional deadline the peer holds
+ *    here, so its connection outlives the deadline even while the peer is still
+ *    unplaced. A refused one disarms nothing.
+ *
+ * The deadline clears on reservation ADMISSION, not on the reservation's own
+ * success: this gate cannot observe the server-side `reserve()` outcome, so a
+ * peer whose admitted reservation is then refused for server capacity keeps its
+ * connection until either side closes it — bounded and mute, so harmless.
+ *
+ * Everything else a control node serves is either libp2p plumbing open to any
+ * connection (identify, ping, hole punching, AutoNAT, the relay protocols) or
+ * members-only — wake, strand-addr, the Optimystic control-DB protocols and FRET —
+ * and the fail-closed layer for all of it is the protocol guard
+ * (`control-protocol-guard.ts`), one seam that wraps every handler at dispatch,
+ * with a protocol nobody classed treated as members-only. The two layers
+ * complement, not duplicate: this connection gate is fail-open over a live DB
+ * read and decides only how long a connection lasts, the protocol guard is
+ * fail-closed and has NO stranger carve-outs — a live invitation admits a
+ * stranger's connection outright for redemption, yet its repo and FRET streams
+ * are still refused.
+ *
+ * ## The bring-up quiet period
+ *
+ * One state is decided BEFORE any of the above, and is not about the remote peer
+ * at all: while this node's control-database bring-up is in flight
+ * (`InboundAdmissionPolicy.bringUpInFlight`), the gate denies BOTH directions —
+ * `denyDialPeer` and `denyInboundEncryptedConnection` — and then opens.
+ *
+ * The invariant it protects is "`ControlDatabase.initialize()` runs while this
+ * node holds ZERO control connections". Building that database is a long chain
+ * of cohort-consulting block probes, and every connected same-party sibling is in
+ * the cohort those probes consult. A sibling that has not yet replicated this
+ * node's `CadrePeer` row correctly refuses them at its own fail-closed per-stream
+ * gate, so ONE connection in this window turns bring-up into
+ * `BlockUnavailableError` and `start()` rejects. Retrying cannot converge: the
+ * condition that would clear the refusal is this node's own row reaching the
+ * sibling, and writing that row needs the database the retry is building. So the
+ * only fix is to not be in the conversation yet — which is what this window is.
+ *
+ * The ordering is arranged so nothing SHOULD open a connection here anyway:
+ * `network.relayAddrs` resolves to a listener that dials nothing and reserves
+ * after bring-up (`relay-addrs.ts`), `controlNetwork.bootstrapNodes` is dialed
+ * after bring-up (`CadreNode.dialControlBootstrapPeers`), and the control-cohort
+ * reconcile pass is scheduled post-start. This window is what makes that a property
+ * rather than an accident of ordering — the live cases it catches are peers the node
+ * remembers from a previous run, which the connection manager auto-dials, and
+ * inbound dials.
+ *
+ * A denial here costs a retry, not a partition: libp2p's connection manager
+ * re-dials on its auto-dial cadence, the reservation supervisor re-drives, and a
+ * denied inbound peer reconnects. The window opens on bring-up FAILURE too (via
+ * `CadreNode.cleanup`), so teardown is never gated.
+ *
+ * ## Fail-open, deliberately
+ *
+ * Outside that window this layer never denies, and it closes a provisional
+ * connection only on a positive answer that the peer is still unplaced. Any
+ * error, missing dependency, timeout or ambiguous state admits the connection
+ * outright (and the reservation), or keeps a provisional connection at its
+ * deadline, and defers to the fail-closed stream gates — a DB hiccup must not
+ * partition a legitimate cadre.
+ */
+
+import debug from 'debug';
+import type { ConnectionGater, PeerId, MultiaddrConnection } from '@libp2p/interface';
+import { withDeadline } from './control-stream.js';
+import { PARTY_RELAY_RESERVATION_TTL_MS } from './relay-server.js';
+import { ADMISSION_DECISION_TIMEOUT_MS, relayedRequestBudgetMs } from './link-budget.js';
+
+const log = debug('sereus:cadre:connection-gater');
+
+/**
+ * How long a provisionally admitted connection lasts before the gate re-asks the
+ * policy and closes it unless the peer is admissible by then (see the module
+ * doc's "Provisional admission").
+ *
+ * {@link relayedRequestBudgetMs}: 28 500 ms at the default declaration. The
+ * longest exchange a stranger legitimately starts on a fresh connection is a
+ * cadre invitation redemption, and the device abandons each member address after
+ * that same budget of its own declared link (`redeemAtMembers`): it covers the
+ * dial, the request, the member's catch-up push, the seat and the redemption. So
+ * a member never cuts an exchange the device is still waiting on, and a
+ * reservation request lands well inside it. This is the value at the default
+ * declaration; `CadreNode` passes the one derived from its OWN declaration.
+ *
+ * NOTE: accepted tradeoff — any stranger can hold a mute connection to any
+ * enrolled member for this long, weighed against letting a device redeem an
+ * invitation at a member that has not yet received its row (maintainer,
+ * 2026-10-07, plan `cadre-invite-redeemable-before-the-row-replicates`). It can
+ * open only stranger-open and transport streams (the protocol guard), and
+ * libp2p's connection-manager limits bound how many it holds. Revisit if stranger
+ * connections show up as load on a member.
+ */
+export const PROVISIONAL_ADMISSION_DEADLINE_MS = relayedRequestBudgetMs();
+
+/**
+ * Bound on the graceful close of an expired provisional connection. It exists
+ * because `AbstractMultiaddrConnection.close()` awaits an `idle`/`drain` event
+ * when the connection still has unsent bytes, and that wait has no timeout of
+ * its own — an unsignalled one never ends. This gate writes nothing to a
+ * stranger, so the wait is not reachable from here; the bound is here so a timer
+ * callback can never hold an unending await.
+ *
+ * Applied through {@link withDeadline}, not `AbortSignal.timeout`: the latter is
+ * not reliably present on React Native/Hermes, which loads this same module.
+ */
+// eslint-disable-next-line no-restricted-syntax -- link-independent: bounds a local connection close inside a timer callback, and expiry aborts the connection instead
+export const PROVISIONAL_ADMISSION_CLOSE_TIMEOUT_MS = 2_000;
+
+/**
+ * Default cap on concurrent relay reservations held by peers the membership
+ * check could not place (see {@link UnauthorizedReservationBudget}). Small on
+ * purpose: it exists for the handful of genuine members whose rows are still
+ * in flight, not as public relay capacity. Overridable per node via
+ * `network.unauthorizedRelayReservationCap` (0 refuses every unauthorized
+ * reservation — the strict pre-seam posture).
+ *
+ * NOTE: this cap shares the relay server's own reservation store, whose
+ * party-run default size is `PARTY_RELAY_MAX_RESERVATIONS` (128,
+ * `relay-server.ts`); unplaced peers may therefore occupy up to this many of
+ * those slots, and the rest are what members and delegates compete for. The two
+ * are kept apart by hand — if this is raised, or an embedder shrinks the store
+ * through `network.relayServerInit.reservations.maxReservations`, keep this one
+ * well under the store's, or a fleet of unplaceable peers can crowd genuine
+ * members out of the server's own store (which the gate cannot override).
+ */
+export const MAX_UNAUTHORIZED_RELAY_RESERVATIONS = 8;
+
+/**
+ * The connection-level outcome of the admission policy:
+ *  - `'admit'` — a peer with a legitimate claim on the connection (member,
+ *    delegate, infra, open stranger window, or any fail-open state); no deadline.
+ *  - `'admit-provisionally'` — a peer the policy cannot place (an outsider, a
+ *    device whose invitation row this node does not hold yet, or a member whose
+ *    row is still replicating — indistinguishable here): admit the connection,
+ *    and at the deadline ask again and close it unless the answer is `'admit'`.
+ *
+ * There is no deny: the gate's only denial is the bring-up quiet period, which
+ * is not a judgement of the peer.
+ */
+export type InboundConnectionVerdict = 'admit' | 'admit-provisionally';
+
+/**
+ * The admission decisions the gate defers to — implemented by
+ * `CadreNode.admitInboundControlConnection` /
+ * `CadreNode.admitControlRelayReservation` (enrollment windows, anchor state,
+ * bootstrap infra, delegate grants, the authorized-member set, the
+ * unauthorized-reservation budget). Kept injectable so the gater's
+ * composition/fail-open behavior is unit-testable without a full node.
+ */
+export interface InboundAdmissionPolicy {
+  /**
+   * Verdict on an inbound encrypted connection from this peer. Asked again at a
+   * provisional connection's deadline, so it must answer from current state.
+   */
+  admitInbound(remotePeerId: string): Promise<InboundConnectionVerdict> | InboundConnectionVerdict;
+  /** Should this peer be granted a circuit-relay reservation slot? */
+  admitRelayReservation(remotePeerId: string): Promise<boolean> | boolean;
+  /**
+   * True while this node's control-database bring-up is in flight — the
+   * {@link createMembershipConnectionGater} quiet period (see the module doc).
+   * Peer-independent, unlike the two decisions above, and SYNCHRONOUS: it is
+   * read on the outbound-dial path, where an await would be a new stall.
+   *
+   * Optional — a policy that omits it is never quiet, which is the right default
+   * for every caller that is not a booting `CadreNode`.
+   */
+  bringUpInFlight?(): boolean;
+}
+
+/**
+ * Bounded budget of concurrent relay reservations for peers the membership
+ * check could not place. `tryAdmit` is the whole surface: an already-admitted
+ * peer refreshes its entry (a reservation refresh never double-counts), a new
+ * peer takes a free slot or is refused. Entries expire after `ttlMs`, which
+ * `CadreNode` passes as the relay server's own resolved `reservationTtl`
+ * (`relay-server.ts`): the server holds an unrefreshed reservation exactly that
+ * long, and a live reserver re-requests (re-hitting the admission hook,
+ * refreshing its entry) well before expiry — so the live entry count tracks the
+ * server's own occupancy without this module reaching into the server's
+ * reservation store. Injectable `now` keeps expiry testable without fake
+ * timers. Authorized members and delegates are never run through this — the
+ * policy admits them before consulting the budget, and `release`s the slot such
+ * a peer took while it was still unplaceable, so the boot-ordering window a
+ * member passes through costs the budget nothing once its row lands.
+ */
+export class UnauthorizedReservationBudget {
+  private readonly admitted = new Map<string, number>();
+
+  constructor(
+    /** Most peers admitted at once — `network.unauthorizedRelayReservationCap`, resolved. */
+    readonly cap: number = MAX_UNAUTHORIZED_RELAY_RESERVATIONS,
+    private readonly ttlMs: number = PARTY_RELAY_RESERVATION_TTL_MS
+  ) {}
+
+  /** Number of live (unexpired at last prune) entries — test/diagnostic surface. */
+  get size(): number {
+    return this.admitted.size;
+  }
+
+  /** Admit (or refresh) `remotePeerId` if a slot is free; false when the cap is spent. */
+  tryAdmit(remotePeerId: string, now: number = Date.now()): boolean {
+    this.prune(now);
+    if (!this.admitted.has(remotePeerId) && this.admitted.size >= this.cap) {
+      log('Unauthorized-reservation budget spent (%d/%d) — refusing %s', this.admitted.size, this.cap, remotePeerId);
+      return false;
+    }
+    this.admitted.set(remotePeerId, now + this.ttlMs);
+    return true;
+  }
+
+  /**
+   * Give back the slot `remotePeerId` holds, if any — called when the peer has
+   * become admissible on its own merits (its membership row replicated, or a
+   * delegate grant landed), so the slot it took during the boot-ordering window
+   * does not stay spent for the remaining TTL on a peer that no longer needs it.
+   */
+  release(remotePeerId: string): void {
+    this.admitted.delete(remotePeerId);
+  }
+
+  private prune(now: number): void {
+    for (const [peerId, expiresAt] of this.admitted) {
+      if (expiresAt <= now) {
+        this.admitted.delete(peerId);
+      }
+    }
+  }
+}
+
+/**
+ * Build the control node's connection gater: the caller-supplied gater (if any)
+ * with membership admission composed onto `denyInboundEncryptedConnection` —
+ * the earliest checkpoint where the remote's authenticated PeerId is known —
+ * and reservation admission composed onto `denyInboundRelayReservation` — the
+ * hook the circuit-relay server consults per RESERVE request (inert on a node
+ * whose relay server is off; libp2p never calls it there).
+ *
+ * `denyDialPeer` is composed too, but ONLY for the bring-up quiet period (see
+ * the module doc): outside that window outbound dials are never gated by
+ * membership — this node decides for itself who to talk to.
+ *
+ * Composition semantics: every hook of `base` is preserved as-is; on the three
+ * composed hooks a deny from the base gater denies, and so does the quiet
+ * period. The admission policy never denies a connection. A policy error — or a
+ * decision slower than `decisionTimeoutMs` — takes the fail-open outcome
+ * (connection admitted outright / reservation admitted, see module doc); the
+ * base gater's verdict is still honored first.
+ *
+ * An `'admit-provisionally'` verdict admits the connection and arms a
+ * `provisionalDeadlineMs` timer against it. When the timer fires the policy is
+ * asked again, and the underlying `MultiaddrConnection` is closed unless it now
+ * answers `'admit'`; a reservation for that peer ADMITTED at the reservation
+ * hook disarms the timer first.
+ *
+ * NOTE: `base` is spread, so a gater passed as a CLASS INSTANCE would lose its
+ * prototype methods; every caller in this repo (and libp2p's own default)
+ * supplies a plain object. If a class-based gater ever shows up, delegate
+ * per-hook instead of spreading.
+ *
+ * Deny timing, as observed by the denied dialer (a quiet-period or base-gater
+ * deny): noise negotiates the muxer in the security handshake's early data, so
+ * the DIALER's upgrade may complete (its `dial()` resolves) before this
+ * receiver-side hook runs. The deny then aborts the receiver's upgrade — the
+ * receiver never registers the connection and never creates its muxer, so no
+ * protocol can ever be negotiated — and the dialer sees its "open" connection
+ * close moments later.
+ *
+ * Control node only: strand cohort nodes legitimately connect cross-party
+ * peers, so cadre membership never gates them. An OPEN strand's node receives
+ * the raw configured gater unchanged; a CLOSED strand's node composes its own
+ * revoked-peer denial onto it instead — a deny-list on positive revocation
+ * evidence, not this module's membership admission — see
+ * `strand-revocation-enforcer.ts`.
+ */
+export function createMembershipConnectionGater(
+  policy: InboundAdmissionPolicy,
+  base?: ConnectionGater,
+  decisionTimeoutMs: number = ADMISSION_DECISION_TIMEOUT_MS,
+  provisionalDeadlineMs: number = PROVISIONAL_ADMISSION_DEADLINE_MS
+): ConnectionGater {
+  const decideInbound = (remotePeerId: string): Promise<InboundConnectionVerdict> => decideWithinDeadline(
+    () => policy.admitInbound(remotePeerId), 'admit', decisionTimeoutMs, `admitInbound(${remotePeerId})`
+  );
+  const provisional = new ProvisionalAdmissions(provisionalDeadlineMs, decideInbound);
+  const quiet = (): boolean => policy.bringUpInFlight?.() ?? false;
+  return {
+    ...base,
+    denyDialPeer: async (peerId: PeerId): Promise<boolean> => {
+      if (await base?.denyDialPeer?.(peerId)) {
+        return true;
+      }
+      if (quiet()) {
+        log('Control-database bring-up in flight — refusing the outbound dial to %s', peerId.toString());
+        return true;
+      }
+      return false;
+    },
+    denyInboundEncryptedConnection: async (peerId: PeerId, maConn: MultiaddrConnection): Promise<boolean> => {
+      if (await base?.denyInboundEncryptedConnection?.(peerId, maConn)) {
+        return true;
+      }
+      if (quiet()) {
+        log('Control-database bring-up in flight — refusing the inbound connection from %s', peerId.toString());
+        return true;
+      }
+      const remotePeerId = peerId.toString();
+      let verdict: InboundConnectionVerdict;
+      try {
+        verdict = await decideInbound(remotePeerId);
+      } catch (error) {
+        log('admitInbound threw for %s — admitting (fail-open; stream gates decide): %o', remotePeerId, error);
+        return false;
+      }
+      if (verdict === 'admit-provisionally') {
+        provisional.arm(remotePeerId, maConn);
+      }
+      return false;
+    },
+    denyInboundRelayReservation: async (peerId: PeerId): Promise<boolean> => {
+      if (await base?.denyInboundRelayReservation?.(peerId)) {
+        return true;
+      }
+      const remotePeerId = peerId.toString();
+      let admitted: boolean;
+      try {
+        admitted = await decideWithinDeadline(
+          () => policy.admitRelayReservation(remotePeerId), true, decisionTimeoutMs, `admitRelayReservation(${remotePeerId})`
+        );
+      } catch (error) {
+        log('admitRelayReservation threw for %s — admitting (fail-open): %o', remotePeerId, error);
+        admitted = true;
+      }
+      if (admitted) {
+        provisional.disarm(remotePeerId);
+      }
+      return !admitted;
+    }
+  };
+}
+
+/**
+ * Run one admission decision under `timeoutMs`, resolving to `fallback` (the
+ * fail-open outcome) if it has not settled in time — see
+ * {@link ADMISSION_DECISION_TIMEOUT_MS} for why an unbounded await is not safe
+ * on the connection hook. The timer is always cleared so a decided call never
+ * holds the event loop open. Exported for the strand revoked-peer gate
+ * (`strand-revocation-enforcer.ts`), which bounds its connection hooks with the
+ * same fail-open contract.
+ */
+export async function decideWithinDeadline<T>(
+  decide: () => Promise<T> | T,
+  fallback: T,
+  timeoutMs: number,
+  what: string
+): Promise<T> {
+  let timer: ReturnType<typeof setTimeout> | undefined;
+  try {
+    return await Promise.race([
+      Promise.resolve().then(decide),
+      new Promise<T>((resolve) => {
+        timer = setTimeout(() => {
+          log('%s exceeded %dms — taking the fail-open outcome (stream gates decide)', what, timeoutMs);
+          resolve(fallback);
+        }, timeoutMs);
+      })
+    ]);
+  } finally {
+    clearTimeout(timer);
+  }
+}
+
+/** One armed provisional deadline: the connection it may close, and its timer. */
+interface ProvisionalAdmission {
+  maConn: MultiaddrConnection;
+  timer: ReturnType<typeof setTimeout>;
+}
+
+/**
+ * The deadlines of `'admit-provisionally'` connections, keyed by remote peerId
+ * (a Set per peer — one peer can hold several in-flight connections). `arm`
+ * starts a timer that re-asks the policy and closes the connection unless it now
+ * admits; `disarm` (called when a reservation for that peer is admitted) cancels
+ * every pending deadline for the peer, one whose question is already out
+ * included. Timers are unref'd so an armed deadline never holds a process open.
+ *
+ * Disarming is final for the connections it cancelled: a peer that reserved
+ * once and then lets its reservation lapse keeps a mute connection (the
+ * fail-closed stream gates refuse it every members-only protocol) until either
+ * side closes it. Bounded — only a peer the reservation policy already admitted
+ * can reach that state, so unplaceable peers are bounded by the budget cap.
+ */
+class ProvisionalAdmissions {
+  private readonly byPeer = new Map<string, Set<ProvisionalAdmission>>();
+
+  constructor(
+    private readonly deadlineMs: number,
+    private readonly decide: (remotePeerId: string) => Promise<InboundConnectionVerdict>
+  ) {}
+
+  arm(remotePeerId: string, maConn: MultiaddrConnection): void {
+    const entry: ProvisionalAdmission = {
+      maConn,
+      timer: setTimeout(() => void this.expire(remotePeerId, entry), this.deadlineMs)
+    };
+    (entry.timer as { unref?: () => void }).unref?.();
+    let entries = this.byPeer.get(remotePeerId);
+    if (!entries) {
+      entries = new Set();
+      this.byPeer.set(remotePeerId, entries);
+    }
+    entries.add(entry);
+    log('Admitted %s provisionally — closing the connection in %dms unless it is admissible by then', remotePeerId, this.deadlineMs);
+  }
+
+  disarm(remotePeerId: string): void {
+    const entries = this.byPeer.get(remotePeerId);
+    if (!entries) {
+      return;
+    }
+    this.byPeer.delete(remotePeerId);
+    for (const entry of entries) {
+      clearTimeout(entry.timer);
+    }
+  }
+
+  private async expire(remotePeerId: string, entry: ProvisionalAdmission): Promise<void> {
+    if (entry.maConn.status !== 'open') {
+      this.release(remotePeerId, entry);
+      return;
+    }
+    const verdict = await this.recheck(remotePeerId);
+    // False when a reservation admitted while the question was out disarmed this deadline.
+    if (!this.release(remotePeerId, entry)) {
+      return;
+    }
+    if (verdict === 'admit') {
+      log('%s is admissible at its provisional deadline — keeping the connection', remotePeerId);
+      return;
+    }
+    log('Provisional admission expired for %s — still unplaced after %dms, closing the connection', remotePeerId, this.deadlineMs);
+    await this.drop(remotePeerId, entry.maConn);
+  }
+
+  /** The policy's answer at the deadline; a throw keeps the connection (fail-open). */
+  private async recheck(remotePeerId: string): Promise<InboundConnectionVerdict> {
+    try {
+      return await this.decide(remotePeerId);
+    } catch (error) {
+      log('admitInbound threw for %s at its provisional deadline — keeping the connection (fail-open): %o', remotePeerId, error);
+      return 'admit';
+    }
+  }
+
+  /** Forget `entry`; false when a `disarm` already took it. */
+  private release(remotePeerId: string, entry: ProvisionalAdmission): boolean {
+    const entries = this.byPeer.get(remotePeerId);
+    if (!entries?.delete(entry)) {
+      return false;
+    }
+    if (entries.size === 0) {
+      this.byPeer.delete(remotePeerId);
+    }
+    return true;
+  }
+
+  /**
+   * End an expired provisional connection: a bounded graceful close, with
+   * `abort()` only as the escalation. The order cannot be reversed — `abort()`
+   * marks the connection `aborted`, and `close()` returns immediately on any
+   * status that is not `open`, so aborting first would make the close a silent
+   * no-op. Closing a connection the peer already dropped returns at once
+   * without throwing, so the fallback stays unreached in the ordinary case.
+   */
+  private async drop(remotePeerId: string, maConn: MultiaddrConnection): Promise<void> {
+    try {
+      await withDeadline(
+        PROVISIONAL_ADMISSION_CLOSE_TIMEOUT_MS,
+        `provisional close for ${remotePeerId}`,
+        (signal) => maConn.close({ signal })
+      );
+      return;
+    } catch (error) {
+      log('Closing the expired provisional connection from %s failed — aborting: %o', remotePeerId, error);
+    }
+    // NOTE: this fallback cannot actually free a WebSocket. `@libp2p/websockets`'
+    // `sendReset()` calls `websocket.close(1006)`, and 1006 is a reserved code
+    // RFC 6455 forbids an endpoint from sending, so `ws` throws and
+    // `AbstractMessageStream.abort()` swallows it — the local end flips to
+    // `aborted` while the socket stays up. Measured against
+    // @libp2p/websockets@10.1.3 (see tickets/blocked/report-libp2p-websockets-abort-close-code.md).
+    // `MultiaddrConnection` exposes no way down to the raw socket, so this is as
+    // far as this layer can go. Revisit — drop the fallback, or go back to a
+    // plain `abort()` — once that transport sends a legal reset code.
+    try {
+      maConn.abort(new Error(`provisional admission expired: still unplaced after ${this.deadlineMs}ms`));
+    } catch (error) {
+      log('Aborting the expired provisional connection from %s threw: %o', remotePeerId, error);
+    }
+  }
+}