@serfab/cadre-core 0.12.0 → 0.13.0

package/src/strand-formation-protocol.ts

@@ -1,692 +1,808 @@
-/**
- * Native cadre-core strand-formation transport.
- *
- * Mirrors the `seed-bootstrap.ts` protocol service: a dedicated
- * libp2p protocol id, length-prefixed JSON frames, and small single-purpose
- * session helpers. It replaces the removed `@serfab/strand-proto` package's transport,
- * carrying the caller's REAL invitation token + disclosure and BOTH parties'
- * real cadre peer addresses end-to-end (no `{ partyId: sessionId }` / `cadre-*.local`
- * placeholders), and validating the responder's result on the initiator side.
- *
- * Provisioning flow (`responderCreates`, 2 messages): the responder provisions (or, for
- * provision-then-record, resolves + records consent against) the strand and returns the
- * result on approval.
- *
- * Cadre-disclosure timing: the responder reveals its own party id + cadre addresses
- * — and, for a closed strand returned via provision-then-record, that strand's
- * membership key — ONLY after the token and disclosure validate; a rejection
- * discloses none of them.
- */
-
-import debug from 'debug';
-import { multiaddr } from '@multiformats/multiaddr';
-import type { Libp2p } from '@libp2p/interface';
-import type { StrandFormationDisclosure } from './types.js';
-import { type ControlStream, writeFrame, withTimeout } from './control-stream.js';
-import { canonicalJson } from './canonical-json.js';
-import { requireEd25519PublicKeyB64 } from './ed25519-key.js';
-import { verifyFormationConsent } from './peer-authorization.js';
-// seed-bootstrap imports neither this protocol nor the manager/solicitation layers,
-// so this import introduces no cycle.
-import { ed25519PublicKeyB64FromPeerId } from './seed-bootstrap.js';
-
-const log = debug('sereus:cadre:formation-proto');
-
-/** Protocol id for the native formation transport (parallel to `/sereus/seed/1.0.0`). */
-export const FORMATION_PROTOCOL = '/sereus/formation/1.0.0';
-
-/**
- * Rejection reason for an invitation that cannot be redeemed at all — unknown, expired, or
- * fully spent. Shared with `StrandFormationManager`, which reports an
- * `InvitationExhaustedError` (an invite spent out from under a redemption already in flight)
- * with this SAME wording on purpose: a joiner that lost the race must be indistinguishable
- * from a non-racing latecomer, so the two sites must never drift apart.
- */
-export const INVALID_TOKEN_REASON = 'Invalid token';
-
-/** Maximum formation message size (1MB). */
-const MAX_FORMATION_MSG_SIZE = 1024 * 1024;
-
-/** Default whole-session timeout (ms). */
-const DEFAULT_SESSION_TIMEOUT_MS = 30_000;
-/** Default per-step (single read/write) timeout (ms). */
-const DEFAULT_STEP_TIMEOUT_MS = 5_000;
-/**
- * Default provisioning budget (ms) — distinct from {@link DEFAULT_STEP_TIMEOUT_MS} because
- * `provisionStrand` is real work (DB writes, and an outbound approval-hook HTTP call when
- * the invite carries a `ValidationUrl`), not a bare wire read/write.
- *
- * Ordering is deliberate — each layer must be able to fail and report before the layer above
- * it gives up:
- *
- *   approval hook (10 s, `formation-approval.ts`)
- *     < responder provisioning (12 s)
- *     < initiator await-response (15 s)
- *     < session (30 s)
- *
- * The 3 s margin between responder provisioning and initiator await-response is the wire
- * latency budget for the result frame to travel back.
- *
- * The last {@link PROVISION_SETTLE_GRACE_MS} of this budget is a settle grace, so the default
- * WORK budget (12 s − 2 s = 10 s) EQUALS the hook's own 10 s: a dead hook races 'Formation
- * approval unavailable, retry' against 'Formation provisioning timed out'. Both are retryable
- * and both leave the invite unspent, so the race is benign.
- */
-const DEFAULT_PROVISION_TIMEOUT_MS = 12_000;
-/**
- * Settle grace (ms), carved OUT of `provisionTimeoutMs` — never added on top, so the budget
- * ladder above is untouched. The listener aborts the provisioning hook when the work budget
- * ({@link splitProvisionBudget}) expires, then waits up to this grace for the work to settle
- * anyway: once the `FormationUsage` insert has been issued nothing can un-spend it (the table
- * is append-only), so a provisioning that lands late is ADOPTED and reported as a real
- * approval rather than lying "timed out" over a spent invite. A hook that observes the
- * abort before writing leaves the invite unspent.
- */
-const PROVISION_SETTLE_GRACE_MS = 2_000;
-/**
- * Wire-latency margin (ms) added to the responder's provisioning budget to get the
- * initiator's await-response budget — the travel time for the result frame to come back
- * after the responder finishes (see {@link DEFAULT_PROVISION_TIMEOUT_MS}).
- * `StrandFormationManager` reuses this constant so a CONFIGURED budget preserves the same
- * margin the defaults do, instead of applying the same number to both sides and racing the
- * initiator's own timeout against the responder's clean rejection reply. The responder's
- * clamp ({@link provisionCeilingMs}) holds the same margin back, so the ordering survives a
- * budget large enough that BOTH roles get clamped.
- *
- * NOTE: 3 s is a fixed assumption about how long the result frame takes to travel back. If
- * formation ever runs over paths where that frame can take longer (a slow circuit relay, a
- * congested mobile link), the margin is too thin and the initiator's timeout races the
- * responder's reply again — raise it, or derive it from an observed round-trip.
- */
-export const PROVISION_RESPONSE_TRAVEL_MARGIN_MS = 3_000;
-/** Default initiator await-response budget (ms); see {@link DEFAULT_PROVISION_TIMEOUT_MS}. */
-const DEFAULT_INITIATOR_PROVISION_TIMEOUT_MS = DEFAULT_PROVISION_TIMEOUT_MS + PROVISION_RESPONSE_TRAVEL_MARGIN_MS;
-/**
- * Default cap on concurrent inbound formation sessions.
- * NOTE: a longer provisioning budget holds a session's slot longer, so a slow hook can
- * exhaust this cap and make the listener reject fresh joiners with 'Too many concurrent
- * formation sessions'. Fine at the current default; revisit if `provisionTimeoutMs` is
- * raised well above its default or this cap is lowered.
- */
-const DEFAULT_MAX_CONCURRENT_SESSIONS = 100;
-
-// ── Roles ────────────────────────────────────────────────────────────────────
-
-export type FormationParty = 'initiator' | 'responder';
-
-// ── Wire messages (length-prefixed JSON frames) ──────────────────────────────
-
-export interface FormationStrandInfo {
-  strandId: string;
-  createdBy: FormationParty;
-}
-
-export interface FormationDbConnectionInfo {
-  endpoint: string;
-  credentialsRef: string;
-}
-
-export interface FormationProvisionResult {
-  strand: FormationStrandInfo;
-  /**
-   * The strand's membership key (closed-strand read-gating secret), delivered to a
-   * validated invitee for provision-then-record formation. Disclosed only AFTER token
-   * + disclosure validation, exactly like the responder identity/cadre — a rejected or
-   * already-used token discloses neither identity nor key. Absent for open strands.
-   */
-  memberPrivateKey?: string;
-  dbConnectionInfo: FormationDbConnectionInfo;
-}
-
-/**
- * Outcome of the responder's provisioning hook.
- *
- * Distinct from a bare {@link FormationProvisionResult} so the hook can REJECT a
- * formation AFTER token + disclosure validation — e.g. a bound invite naming a host
- * strand this responder has not yet converged on, or a concurrent-redemption collision.
- * Without this channel such cases threw inside provisioning, the stream closed with no
- * result frame, and the initiator saw a read-error/timeout instead of a clean
- * `approved: false`. A rejection still discloses NO responder identity/cadre.
- */
-export type ResponderProvisionOutcome =
-  | { approved: true; result: FormationProvisionResult }
-  | { approved: false; reason: string };
-
-/** Initiator → Responder: carries the real token + disclosure + initiator cadre. */
-export interface FormationContactMessage {
-  /** The real invitation token. */
-  token: string;
-  /** Initiator's member key (peer id) — human-facing identity in logs and `FormStrandResult.memberKey`. */
-  partyId: string;
-  /**
-   * Initiator's base64url ed25519 public key — the key behind `partyId`, written to
-   * `FormationUsage.PeerKey` and covered by BOTH signed digests.
-   */
-  peerKey: string;
-  /** Joiner-minted single-use nonce for THIS redemption; both signed digests cover it. */
-  usageStampId: string;
-  /**
-   * The joiner's signature over the `'consent'` digest (see `formationConsentMessage`
-   * in control-database.ts) — proof the named peer agreed to this redemption.
-   */
-  peerSignature: string;
-  /** The real disclosure, carried verbatim. */
-  disclosure: StrandFormationDisclosure;
-  /** Initiator's real multiaddrs. */
-  cadrePeerAddrs: string[];
-}
-
-/** Responder → Initiator: responder identity/cadre disclosed only after validation. */
-export interface FormationResultMessage {
-  approved: boolean;
-  /** Present iff `approved === false`. */
-  reason?: string;
-  /** Responder's real party id (disclosed only after validation). */
-  partyId?: string;
-  /** Responder's real multiaddrs (omitted on rejection). */
-  cadrePeerAddrs?: string[];
-  /** The provisioned strand/db result (always present on approval). */
-  provisionResult?: FormationProvisionResult;
-}
-
-// ── Stream framing helpers ───────────────────────────────────────────────────
-
-/**
- * Reads length-prefixed JSON frames from a libp2p stream one frame at a time.
- *
- * Unlike seed delivery (which reads a stream to EOF then parses once), formation
- * is request/response on a single live stream, so reading must stop at each frame
- * boundary. The reader buffers across chunk boundaries and retains any overflow
- * (a following frame coalesced into the same chunk) for the next read.
- */
-class FrameReader {
-  private readonly iterator: AsyncIterator<Uint8Array>;
-  private readonly maxLength: number;
-  private buffer = new Uint8Array(0);
-
-  constructor(stream: AsyncIterable<Uint8Array>, maxLength = MAX_FORMATION_MSG_SIZE) {
-    this.iterator = stream[Symbol.asyncIterator]();
-    this.maxLength = maxLength;
-  }
-
-  private append(chunk: Uint8Array): void {
-    const next = new Uint8Array(this.buffer.length + chunk.length);
-    next.set(this.buffer, 0);
-    next.set(chunk, this.buffer.length);
-    this.buffer = next;
-  }
-
-  /** Pull chunks until the buffer holds ≥ `need` bytes; false if the stream ended first. */
-  private async fill(need: number): Promise<boolean> {
-    while (this.buffer.length < need) {
-      const { value, done } = await this.iterator.next();
-      if (done) return false;
-      const bytes = value instanceof Uint8Array
-        ? value
-        : (value as { subarray(): Uint8Array }).subarray();
-      this.append(bytes);
-    }
-    return true;
-  }
-
-  async read<T>(): Promise<T> {
-    if (!(await this.fill(4))) {
-      throw new Error('Formation stream closed before length prefix');
-    }
-    const length = new DataView(this.buffer.buffer, this.buffer.byteOffset, this.buffer.byteLength).getUint32(0, false);
-    if (length > this.maxLength) {
-      throw new Error(`Formation frame declares length ${length} exceeding max ${this.maxLength}`);
-    }
-    if (!(await this.fill(4 + length))) {
-      throw new Error(`Formation frame declares length ${length} but stream ended early`);
-    }
-    const json = new TextDecoder().decode(this.buffer.subarray(4, 4 + length));
-    this.buffer = this.buffer.subarray(4 + length);
-    return JSON.parse(json) as T;
-  }
-}
-
-// ── Result validation ────────────────────────────────────────────────────────
-
-/** Matches the placeholder cadre addrs (`cadre-a-1.local`, …) of the removed strand-proto transport. */
-const PLACEHOLDER_CADRE_ADDR = /^cadre-[ab]-\d+\.local$/;
-
-/**
- * Structural check on a `responderCreates` result: the responder must have approved,
- * disclosed a non-empty real identity + cadre (not the legacy `cadre-*.local`
- * placeholders), and returned a strand vouched for by the responder party with a real id.
- *
- * `createdBy: 'responder'` means "the responder party vouches for / returns this strand."
- * Under provision-then-record that strand was minted owner-signed earlier (not created
- * in-session), but it is still the responder party returning its own strand, so the marker
- * — and this structural validator — stay unchanged.
- *
- * The behavioral floor for the initiator: a responder that returns an arbitrary/empty
- * `strandId` or omits its disclosed identity is rejected, not silently accepted.
- */
-export function isValidResponderCreatesResult(response: FormationResultMessage): boolean {
-  if (!response.approved) return false;
-  if (!response.partyId) return false;
-  const addrs = response.cadrePeerAddrs;
-  if (!addrs || addrs.length === 0) return false;
-  if (addrs.some((a) => PLACEHOLDER_CADRE_ADDR.test(a))) return false;
-  const provision = response.provisionResult;
-  if (!provision?.strand?.strandId) return false;
-  if (provision.strand.createdBy !== 'responder') return false;
-  return true;
-}
-
-/**
- * Ceiling for a resolved provisioning budget.
- *
- * The session budget also has to cover the wire step that PRECEDES provisioning (the
- * responder's contact read, the initiator's dial-connect), so a whole step is held back
- * rather than triggering only on a literal overrun.
- *
- * `reserveMs` is the EXTRA room only the responder holds back, so its own budget still
- * lands strictly before the initiator's larger await-response budget even when both are
- * clamped — without it, any configured budget at or above `sessionTimeoutMs -
- * stepTimeoutMs` clamps both roles onto the same number and the responder's clean
- * rejection races the initiator's timeout again. Capped at half the remaining room (like
- * {@link splitProvisionBudget}) so a small session config still spends most of it working.
- */
-function provisionCeilingMs(sessionTimeoutMs: number, stepTimeoutMs: number, reserveMs: number): number {
-  const roomMs = sessionTimeoutMs - stepTimeoutMs;
-  return Math.max(1, roomMs - Math.min(reserveMs, Math.floor(roomMs / 2)));
-}
-
-/**
- * Resolve a caller-supplied provisioning budget: `0`/negative means "unset" (use
- * `defaultMs`). If the result would let provisioning outlive the session — no result
- * frame is ever sent, exactly the failure this budget exists to prevent — clamp it to
- * {@link provisionCeilingMs} and log a warning.
- */
-function resolveProvisionTimeoutMs(
-  configured: number | undefined,
-  defaultMs: number,
-  sessionTimeoutMs: number,
-  stepTimeoutMs: number,
-  role: string,
-  reserveMs = 0
-): number {
-  const requested = configured && configured > 0 ? configured : defaultMs;
-  const ceilingMs = provisionCeilingMs(sessionTimeoutMs, stepTimeoutMs, reserveMs);
-  if (requested >= ceilingMs) {
-    log(
-      '%s provisionTimeoutMs %dms leaves no room under sessionTimeoutMs %dms (step %dms, reserve %dms); clamping to %dms',
-      role, requested, sessionTimeoutMs, stepTimeoutMs, reserveMs, ceilingMs
-    );
-    return ceilingMs;
-  }
-  return requested;
-}
-
-/**
- * Split a resolved provisioning budget into the WORK budget and the trailing settle grace.
- *
- * The grace is carved OUT of the budget (see {@link PROVISION_SETTLE_GRACE_MS}), and capped at
- * half of it so a small configured budget still spends at least half its time doing work.
- */
-function splitProvisionBudget(provisionTimeoutMs: number): { workMs: number; graceMs: number } {
-  const graceMs = Math.min(PROVISION_SETTLE_GRACE_MS, Math.floor(provisionTimeoutMs / 2));
-  return { workMs: provisionTimeoutMs - graceMs, graceMs };
-}
-
-// ── Responder (listener) ─────────────────────────────────────────────────────
-
-export interface FormationListenerOptions {
-  /** Validate the invitation token; returns whether it is valid. */
-  validateToken(token: string): Promise<{ valid: boolean }>;
-  /** Validate the initiator's disclosure with the REAL token + disclosure. */
-  validateDisclosure(token: string, disclosure: StrandFormationDisclosure): Promise<boolean>;
-  /**
-   * Provision (or, for provision-then-record, resolve + record consent against) the
-   * strand (`responderCreates`) for the given initiator. The WHOLE contact message is
-   * threaded in: the hook needs the REAL token to map the bound host strand, and the
-   * joiner's `peerKey`/`usageStampId`/`peerSignature` to write the consent columns.
-   *
-   * Returns a {@link ResponderProvisionOutcome}: the hook may REJECT post-validation
-   * (e.g. an unconverged host strand) and the listener turns that into a clean,
-   * non-disclosing `approved: false` reply rather than dropping the result frame.
-   *
-   * `signal` is aborted when the listener's work budget expires: a hook that observes it
-   * BEFORE writing abandons the redemption and leaves the invite unspent. Optional so
-   * signal-unaware hooks (tests, mocks) stay assignable.
-   */
-  provisionStrand(contact: FormationContactMessage, signal?: AbortSignal): Promise<ResponderProvisionOutcome>;
-  /** Responder identity, disclosed only AFTER token + disclosure validation passes. */
-  getResponderIdentity(): { partyId: string; cadrePeerAddrs: string[] };
-  sessionTimeoutMs?: number;
-  stepTimeoutMs?: number;
-  /**
-   * Budget for the `provisionStrand` hook call — distinct from `stepTimeoutMs` because
-   * provisioning is real work, not a bare wire read/write. Default 12 s; `0`/unset uses
-   * the default; a value that would outlive the session is clamped ({@link
-   * provisionCeilingMs}). See {@link DEFAULT_PROVISION_TIMEOUT_MS} for the full ordering
-   * rationale.
-   */
-  provisionTimeoutMs?: number;
-  maxConcurrentSessions?: number;
-}
-
-/**
- * Responder side of the native formation protocol. Registers a libp2p handler and
- * drives each inbound stream through token + disclosure validation, provisioning,
- * and result delivery — enforcing the cadre-disclosure timing rule (responder cadre
- * is revealed only after validation; rejections disclose nothing).
- */
-export class FormationListener {
-  private readonly options: FormationListenerOptions;
-  private readonly sessionTimeoutMs: number;
-  private readonly stepTimeoutMs: number;
-  private readonly provisionWorkMs: number;
-  private readonly provisionGraceMs: number;
-  private readonly maxConcurrentSessions: number;
-  private readonly registered = new Set<Libp2p>();
-  private activeSessions = 0;
-  private sessionCounter = 0;
-
-  constructor(options: FormationListenerOptions) {
-    this.options = options;
-    this.sessionTimeoutMs = options.sessionTimeoutMs ?? DEFAULT_SESSION_TIMEOUT_MS;
-    this.stepTimeoutMs = options.stepTimeoutMs ?? DEFAULT_STEP_TIMEOUT_MS;
-    const split = splitProvisionBudget(resolveProvisionTimeoutMs(
-      options.provisionTimeoutMs, DEFAULT_PROVISION_TIMEOUT_MS,
-      this.sessionTimeoutMs, this.stepTimeoutMs, 'FormationListener',
-      PROVISION_RESPONSE_TRAVEL_MARGIN_MS
-    ));
-    this.provisionWorkMs = split.workMs;
-    this.provisionGraceMs = split.graceMs;
-    this.maxConcurrentSessions = options.maxConcurrentSessions ?? DEFAULT_MAX_CONCURRENT_SESSIONS;
-  }
-
-  /** Number of in-flight inbound sessions. */
-  get activeCount(): number {
-    return this.activeSessions;
-  }
-
-  register(node: Libp2p, protocolId: string = FORMATION_PROTOCOL): void {
-    if (this.registered.has(node)) {
-      log('node already registered');
-      return;
-    }
-    void node.handle(protocolId, async (rawStream: unknown, _connection: unknown) => {
-      await this.handleStream(rawStream as ControlStream);
-    });
-    this.registered.add(node);
-    log('formation listener registered (%s)', protocolId);
-  }
-
-  unregister(node: Libp2p, protocolId: string = FORMATION_PROTOCOL): void {
-    if (!this.registered.has(node)) return;
-    void node.unhandle(protocolId);
-    this.registered.delete(node);
-    log('formation listener unregistered (%s)', protocolId);
-  }
-
-  private async handleStream(stream: ControlStream): Promise<void> {
-    if (this.activeSessions >= this.maxConcurrentSessions) {
-      const rejection: FormationResultMessage = { approved: false, reason: 'Too many concurrent formation sessions' };
-      try { writeFrame(stream, rejection); } catch { /* best effort */ }
-      try { await stream.close(); } catch { /* ignore */ }
-      return;
-    }
-    const id = ++this.sessionCounter;
-    this.activeSessions++;
-    try {
-      await withTimeout(this.sessionTimeoutMs, `Formation session#${id}`, () => this.runSession(id, stream));
-    } catch (err) {
-      log('formation session #%d failed: %o', id, err);
-    } finally {
-      this.activeSessions--;
-      try { await stream.close(); } catch { /* ignore */ }
-    }
-  }
-
-  /**
-   * Run the provisioning hook under its WORK budget — the configured provisioning budget minus
-   * the trailing settle grace ({@link splitProvisionBudget}, {@link PROVISION_SETTLE_GRACE_MS}).
-   *
-   * When the work budget expires the hook's signal is aborted, then
-   * {@link settleWithinGrace} waits out the grace for the aborted work to settle anyway.
-   * Returns `undefined` for "timed out" — which the caller turns into a retryable rejection
-   * frame — and rethrows every other failure so it reaches the internal-error path.
-   *
-   * Deliberately not `withDeadline`: that exposes neither the timed-out flag nor the signal
-   * outside its `op`, and both are needed here.
-   */
-  private async provision(id: number, contact: FormationContactMessage): Promise<ResponderProvisionOutcome | undefined> {
-    const controller = new AbortController();
-    let timedOut = false;
-    const pending = this.options.provisionStrand(contact, controller.signal);
-    try {
-      return await withTimeout(
-        this.provisionWorkMs,
-        `Formation provisioning#${id}`,
-        () => pending,
-        () => { timedOut = true; controller.abort(); }
-      );
-    } catch (err) {
-      if (!timedOut) throw err;
-      log(
-        'formation session #%d provisioning work budget expired after %dms; aborted, settling up to %dms',
-        id, this.provisionWorkMs, this.provisionGraceMs
-      );
-      return await this.settleWithinGrace(id, pending);
-    }
-  }
-
-  /**
-   * Wait out the settle grace for an ABORTED provisioning, adopting a late success.
-   *
-   * A hook that observed the abort before issuing its `FormationUsage` insert rejects (with
-   * `FormationAbortedError`) and leaves the invite unspent → reported as a timeout. A hook
-   * that had already written cannot be un-written (the table is append-only), so if it lands
-   * inside the grace its outcome is ADOPTED — the joiner is told the truth (approved, or the
-   * hook's own rejection reason) rather than "timed out" over a spent invite.
-   *
-   * NOTE: one window stays open — a work budget that expires after the `FormationUsage`
-   * insert was issued AND a commit that outlasts the grace. The joiner is told 'timed out'
-   * while its one-time invite is in fact spent, and since every retry mints a fresh keypair
-   * no recovery path can match it. Nothing can un-spend an append-only row, so the late-settle
-   * logging below is the observability for it; widen the grace only if it is seen in the wild.
-   */
-  private async settleWithinGrace(
-    id: number,
-    pending: Promise<ResponderProvisionOutcome>
-  ): Promise<ResponderProvisionOutcome | undefined> {
-    let stillPending = false;
-    try {
-      const outcome = await withTimeout(
-        this.provisionGraceMs,
-        `Formation settle#${id}`,
-        () => pending,
-        () => { stillPending = true; }
-      );
-      log('formation session #%d provisioning settled within its grace: approved=%s — adopting', id, outcome.approved);
-      return outcome;
-    } catch (err) {
-      if (!stillPending) {
-        log('formation session #%d provisioning failed after its abort: %o', id, err);
-        return undefined;
-      }
-      log('formation session #%d provisioning still pending after its %dms grace', id, this.provisionGraceMs);
-      // Log how it eventually settled so a late failure is not swallowed by the two
-      // abandoned `withTimeout` calls.
-      void pending.then(
-        (late) => log('formation session #%d provisioning settled after its grace: approved=%s', id, late.approved),
-        (err2) => log('formation session #%d provisioning failed after its grace: %o', id, err2)
-      );
-      return undefined;
-    }
-  }
-
-  /**
-   * Joiner-consent pre-check — the legibility twin of the schema's
-   * `FormationUsage.PeerConsented` CHECK, exactly as `verifyFormationApproval` is for the
-   * approver's vouch. The database constraint stays the authority; this turns a bad consent
-   * into a clean 'Invalid joiner consent' rejection instead of an opaque CHECK failure at
-   * commit. Runs BEFORE token/disclosure validation: it is the cheapest check (pure local
-   * crypto, no DB read) and a rejection discloses nothing.
-   *
-   * Also pins `partyId` to `peerKey`: an ed25519 peer id is an identity multihash of the
-   * very key, and this is the ONLY layer that can check the match — SQL cannot unwrap a
-   * multihash, and `partyId` itself is never stored.
-   */
-  private isJoinerConsentValid(id: number, contact: FormationContactMessage): boolean {
-    try {
-      requireEd25519PublicKeyB64(contact.peerKey, 'joiner peer key');
-    } catch {
-      // Deliberately not rethrown or echoed: the helper's error message embeds the whole
-      // rejected value, and this one arrived from a remote peer (the cap-the-echo case its
-      // own NOTE warns about). Log without the raw value instead.
-      log('formation session #%d: joiner peer key is not a base64url ed25519 public key', id);
-      return false;
-    }
-    const embedded = ed25519PublicKeyB64FromPeerId(contact.partyId);
-    if (!embedded || embedded !== contact.peerKey) {
-      log('formation session #%d: partyId does not embed the presented peer key', id);
-      return false;
-    }
-    const consentOk = verifyFormationConsent({
-      token: contact.token,
-      usageStampId: contact.usageStampId,
-      peerKey: contact.peerKey,
-      disclosure: canonicalJson(contact.disclosure),
-      peerSig: contact.peerSignature
-    });
-    if (!consentOk) {
-      log('formation session #%d: joiner consent signature failed verification', id);
-      return false;
-    }
-    return true;
-  }
-
-  private async runSession(id: number, stream: ControlStream): Promise<void> {
-    // Track whether ANY frame has been written so the catch below can convert an
-    // unexpected internal error into a non-disclosing rejection ONLY when nothing has
-    // gone out yet. This closes the "stream closed with no result frame" class of bug.
-    let wroteFrame = false;
-    const send = (msg: FormationResultMessage): void => {
-      writeFrame(stream, msg);
-      wroteFrame = true;
-    };
-
-    try {
-      const reader = new FrameReader(stream);
-      const contact = await withTimeout(this.stepTimeoutMs, `Formation await-contact#${id}`, () => reader.read<FormationContactMessage>());
-      log('formation session #%d contact: token=%s party=%s', id, contact.token, contact.partyId);
-
-      if (!this.isJoinerConsentValid(id, contact)) {
-        send({ approved: false, reason: 'Invalid joiner consent' });
-        return;
-      }
-
-      const tokenResult = await this.options.validateToken(contact.token);
-      if (!tokenResult.valid) {
-        send({ approved: false, reason: INVALID_TOKEN_REASON });
-        return;
-      }
-
-      const disclosureOk = await this.options.validateDisclosure(contact.token, contact.disclosure);
-      if (!disclosureOk) {
-        send({ approved: false, reason: 'Invalid disclosure' });
-        return;
-      }
-
-      const outcome = await this.provision(id, contact);
-      if (!outcome) {
-        // Reported as its own retryable reason rather than falling through to the generic
-        // catch below, which would report the misleading 'Internal formation error'.
-        send({ approved: false, reason: 'Formation provisioning timed out' });
-        return;
-      }
-      if (!outcome.approved) {
-        // A post-validation rejection still discloses NEITHER identity NOR cadre,
-        // exactly like the token/disclosure rejections above.
-        send({ approved: false, reason: outcome.reason });
-        return;
-      }
-      // Validation + provisioning passed → safe to disclose responder identity/cadre.
-      const identity = this.options.getResponderIdentity();
-      send({
-        approved: true,
-        partyId: identity.partyId,
-        cadrePeerAddrs: identity.cadrePeerAddrs,
-        provisionResult: outcome.result
-      });
-    } catch (err) {
-      // Defense-in-depth: if an unexpected internal error escapes BEFORE any frame was
-      // written (e.g. a future provisionStrand-hook bug), convert it into a non-disclosing
-      // protocol rejection so the initiator sees a clean error instead of a read-error/
-      // timeout. If a frame was already sent, leave the stream as-is and let handleStream
-      // log + close. Re-throw either way so the failure is still recorded — this is a
-      // deliberate, logged conversion, not silent.
-      if (!wroteFrame) {
-        const internalError: FormationResultMessage = { approved: false, reason: 'Internal formation error' };
-        try { writeFrame(stream, internalError); } catch { /* stream already broken */ }
-      }
-      throw err;
-    }
-  }
-}
-
-// ── Initiator (dialer) ───────────────────────────────────────────────────────
-
-export interface FormationDialOptions {
-  /** The contact message to send (real token, partyId, disclosure, initiator cadre). */
-  contact: FormationContactMessage;
-  /** Responder multiaddrs to dial. */
-  responderAddrs: string[];
-  /** Validate the responder's result; a false return aborts the formation. */
-  validateResponse(response: FormationResultMessage): Promise<boolean>;
-  sessionTimeoutMs?: number;
-  stepTimeoutMs?: number;
-  /**
-   * Budget for the `await-response` read only (the responder's provisioning + reply travel
-   * time) — `dial-connect` keeps using `stepTimeoutMs`. Default 15 s; `0`/unset uses the
-   * default. See {@link DEFAULT_PROVISION_TIMEOUT_MS} for the full ordering rationale.
-   */
-  provisionTimeoutMs?: number;
-  protocolId?: string;
-}
-
-/**
- * Initiator side of the native formation protocol. Dials the responder, sends the
- * contact (carrying the real disclosure/token/cadre), validates the responder's
- * result, and returns the strand the responder provisioned.
- */
-export async function dialFormation(node: Libp2p, options: FormationDialOptions): Promise<FormationProvisionResult> {
-  if (options.responderAddrs.length === 0) {
-    throw new Error('No responder addresses available for formation');
-  }
-  const protocolId = options.protocolId ?? FORMATION_PROTOCOL;
-  const sessionTimeoutMs = options.sessionTimeoutMs ?? DEFAULT_SESSION_TIMEOUT_MS;
-  const stepTimeoutMs = options.stepTimeoutMs ?? DEFAULT_STEP_TIMEOUT_MS;
-  const provisionTimeoutMs = resolveProvisionTimeoutMs(
-    options.provisionTimeoutMs, DEFAULT_INITIATOR_PROVISION_TIMEOUT_MS,
-    sessionTimeoutMs, stepTimeoutMs, 'dialFormation'
-  );
-  const addr = multiaddr(options.responderAddrs[0]);
-
-  return withTimeout(sessionTimeoutMs, 'Formation dial', async () => {
-    const rawStream = await withTimeout(stepTimeoutMs, 'Formation dial-connect', async () => node.dialProtocol(addr, protocolId));
-    const stream = rawStream as unknown as ControlStream;
-    try {
-      const reader = new FrameReader(stream);
-      writeFrame(stream, options.contact);
-
-      const response = await withTimeout(provisionTimeoutMs, 'Formation await-response', () => reader.read<FormationResultMessage>());
-      if (!response.approved) {
-        throw new Error(`Formation rejected: ${response.reason ?? 'no reason provided'}`);
-      }
-      const ok = await options.validateResponse(response);
-      if (!ok) throw new Error('Responder result failed validation');
-
-      if (!response.provisionResult) throw new Error('Missing provision result for responderCreates mode');
-      return response.provisionResult;
-    } finally {
-      try { await stream.close(); } catch { /* ignore */ }
-    }
-  });
-}
+/**
+ * Native cadre-core strand-formation transport.
+ *
+ * Mirrors the `seed-bootstrap.ts` protocol service: a dedicated
+ * libp2p protocol id, length-prefixed JSON frames, and small single-purpose
+ * session helpers. It replaces the removed `@serfab/strand-proto` package's transport,
+ * carrying the caller's REAL invitation token + disclosure and BOTH parties'
+ * real cadre peer addresses end-to-end (no `{ partyId: sessionId }` / `cadre-*.local`
+ * placeholders), and validating the responder's result on the initiator side.
+ *
+ * Provisioning flow (`responderCreates`, 2 messages): the responder provisions (or, for
+ * provision-then-record, resolves + records consent against) the strand and returns the
+ * result on approval.
+ *
+ * Cadre-disclosure timing: the responder reveals its own party id + cadre addresses
+ * — and, for a closed strand returned via provision-then-record, that strand's
+ * membership key — ONLY after the token and disclosure validate; a rejection
+ * discloses none of them.
+ */
+
+import debug from 'debug';
+import { multiaddr } from '@multiformats/multiaddr';
+import type { Libp2p } from '@libp2p/interface';
+import type { StrandFormationDisclosure } from './types.js';
+import { type ControlStream, writeFrame, withTimeout } from './control-stream.js';
+import { canonicalJson } from './canonical-json.js';
+import { requireEd25519PublicKeyB64 } from './ed25519-key.js';
+import { verifyFormationConsent } from './peer-authorization.js';
+// seed-bootstrap imports neither this protocol nor the manager/solicitation layers,
+// so this import introduces no cycle.
+import { ed25519PublicKeyB64FromPeerId } from './seed-bootstrap.js';
+
+const log = debug('sereus:cadre:formation-proto');
+
+/** Protocol id for the native formation transport (parallel to `/sereus/seed/1.0.0`). */
+export const FORMATION_PROTOCOL = '/sereus/formation/1.0.0';
+
+/**
+ * Rejection reason for an invitation that cannot be redeemed at all — unknown, expired, or
+ * fully spent. Shared with `StrandFormationManager`, which reports an
+ * `InvitationExhaustedError` (an invite spent out from under a redemption already in flight)
+ * with this SAME wording on purpose: a joiner that lost the race must be indistinguishable
+ * from a non-racing latecomer, so the two sites must never drift apart.
+ */
+export const INVALID_TOKEN_REASON = 'Invalid token';
+
+/** Maximum formation message size (1MB). */
+const MAX_FORMATION_MSG_SIZE = 1024 * 1024;
+
+/** Default whole-session timeout (ms). */
+const DEFAULT_SESSION_TIMEOUT_MS = 30_000;
+/** Default per-step (single read/write) timeout (ms). */
+const DEFAULT_STEP_TIMEOUT_MS = 5_000;
+/**
+ * Default provisioning budget (ms) — distinct from {@link DEFAULT_STEP_TIMEOUT_MS} because
+ * `provisionStrand` is real work (DB writes, and an outbound approval-hook HTTP call when
+ * the invite carries a `ValidationUrl`), not a bare wire read/write.
+ *
+ * Ordering is deliberate — each layer must be able to fail and report before the layer above
+ * it gives up:
+ *
+ *   approval hook (10 s, `formation-approval.ts`)
+ *     < responder provisioning (12 s)
+ *     < initiator await-response (15 s)
+ *     < session (30 s)
+ *
+ * The 3 s margin between responder provisioning and initiator await-response is the wire
+ * latency budget for the result frame to travel back.
+ *
+ * The last {@link PROVISION_SETTLE_GRACE_MS} of this budget is a settle grace, so the default
+ * WORK budget (12 s − 2 s = 10 s) EQUALS the hook's own 10 s: a dead hook races 'Formation
+ * approval unavailable, retry' against 'Formation provisioning timed out'. Both are retryable
+ * and both leave the invite unspent, so the race is benign.
+ */
+const DEFAULT_PROVISION_TIMEOUT_MS = 12_000;
+/**
+ * Settle grace (ms), carved OUT of `provisionTimeoutMs` — never added on top, so the budget
+ * ladder above is untouched. The listener aborts the provisioning hook when the work budget
+ * ({@link splitProvisionBudget}) expires, then waits up to this grace for the work to settle
+ * anyway: once the `FormationUsage` insert has been issued nothing can un-spend it (the table
+ * is append-only), so a provisioning that lands late is ADOPTED and reported as a real
+ * approval rather than lying "timed out" over a spent invite. A hook that observes the
+ * abort before writing leaves the invite unspent.
+ */
+const PROVISION_SETTLE_GRACE_MS = 2_000;
+/**
+ * Wire-latency margin (ms) added to the responder's provisioning budget to get the
+ * initiator's await-response budget — the travel time for the result frame to come back
+ * after the responder finishes (see {@link DEFAULT_PROVISION_TIMEOUT_MS}).
+ * `StrandFormationManager` reuses this constant so a CONFIGURED budget preserves the same
+ * margin the defaults do, instead of applying the same number to both sides and racing the
+ * initiator's own timeout against the responder's clean rejection reply. The responder's
+ * clamp ({@link provisionCeilingMs}) holds the same margin back, so the ordering survives a
+ * budget large enough that BOTH roles get clamped.
+ *
+ * NOTE: 3 s is a fixed assumption about how long the result frame takes to travel back. If
+ * formation ever runs over paths where that frame can take longer (a slow circuit relay, a
+ * congested mobile link), the margin is too thin and the initiator's timeout races the
+ * responder's reply again — raise it, or derive it from an observed round-trip.
+ */
+export const PROVISION_RESPONSE_TRAVEL_MARGIN_MS = 3_000;
+/** Default initiator await-response budget (ms); see {@link DEFAULT_PROVISION_TIMEOUT_MS}. */
+const DEFAULT_INITIATOR_PROVISION_TIMEOUT_MS = DEFAULT_PROVISION_TIMEOUT_MS + PROVISION_RESPONSE_TRAVEL_MARGIN_MS;
+/**
+ * Default cap on concurrent inbound formation sessions.
+ * NOTE: a longer provisioning budget holds a session's slot longer, so a slow hook can
+ * exhaust this cap and make the listener reject fresh joiners with 'Too many concurrent
+ * formation sessions'. Fine at the current default; revisit if `provisionTimeoutMs` is
+ * raised well above its default or this cap is lowered.
+ */
+const DEFAULT_MAX_CONCURRENT_SESSIONS = 100;
+
+/**
+ * Cap on the strand-network addresses a formation result carries
+ * ({@link FormationResultMessage.strandAddrs}). A node's own strand multiaddrs are a
+ * handful of entries (one per listen transport, plus a circuit-relay reservation), so
+ * 16 is generous for the honest case while keeping the result frame bounded against a
+ * peer that would pad it — either direction, since both sides run the list through
+ * {@link sanitizeStrandAddrs}. Exported because the same bound has to hold on the
+ * initiator's stored copy: `CadreNode` accumulates across repeat formations against one
+ * strand, so it caps the accumulation rather than only each arriving list.
+ */
+export const MAX_STRAND_ADDRS = 16;
+
+// ── Roles ────────────────────────────────────────────────────────────────────
+
+export type FormationParty = 'initiator' | 'responder';
+
+// ── Wire messages (length-prefixed JSON frames) ──────────────────────────────
+
+export interface FormationStrandInfo {
+  strandId: string;
+  createdBy: FormationParty;
+}
+
+export interface FormationDbConnectionInfo {
+  endpoint: string;
+  credentialsRef: string;
+}
+
+export interface FormationProvisionResult {
+  strand: FormationStrandInfo;
+  /**
+   * The strand's membership key (closed-strand read-gating secret), delivered to a
+   * validated invitee for provision-then-record formation. Disclosed only AFTER token
+   * + disclosure validation, exactly like the responder identity/cadre — a rejected or
+   * already-used token discloses neither identity nor key. Absent for open strands.
+   */
+  memberPrivateKey?: string;
+  dbConnectionInfo: FormationDbConnectionInfo;
+}
+
+/**
+ * Outcome of the responder's provisioning hook.
+ *
+ * Distinct from a bare {@link FormationProvisionResult} so the hook can REJECT a
+ * formation AFTER token + disclosure validation — e.g. a bound invite naming a host
+ * strand this responder has not yet converged on, or a concurrent-redemption collision.
+ * Without this channel such cases threw inside provisioning, the stream closed with no
+ * result frame, and the initiator saw a read-error/timeout instead of a clean
+ * `approved: false`. A rejection still discloses NO responder identity/cadre.
+ */
+export type ResponderProvisionOutcome =
+  | { approved: true; result: FormationProvisionResult }
+  | { approved: false; reason: string };
+
+/** Initiator → Responder: carries the real token + disclosure + initiator cadre. */
+export interface FormationContactMessage {
+  /** The real invitation token. */
+  token: string;
+  /** Initiator's member key (peer id) — human-facing identity in logs and `FormStrandResult.memberKey`. */
+  partyId: string;
+  /**
+   * Initiator's base64url ed25519 public key — the key behind `partyId`, written to
+   * `FormationUsage.PeerKey` and covered by BOTH signed digests.
+   */
+  peerKey: string;
+  /** Joiner-minted single-use nonce for THIS redemption; both signed digests cover it. */
+  usageStampId: string;
+  /**
+   * The joiner's signature over the `'consent'` digest (see `formationConsentMessage`
+   * in control-database.ts) — proof the named peer agreed to this redemption.
+   */
+  peerSignature: string;
+  /** The real disclosure, carried verbatim. */
+  disclosure: StrandFormationDisclosure;
+  /** Initiator's real multiaddrs. */
+  cadrePeerAddrs: string[];
+}
+
+/** Responder → Initiator: responder identity/cadre disclosed only after validation. */
+export interface FormationResultMessage {
+  approved: boolean;
+  /** Present iff `approved === false`. */
+  reason?: string;
+  /** Responder's real party id (disclosed only after validation). */
+  partyId?: string;
+  /** Responder's real multiaddrs (omitted on rejection). */
+  cadrePeerAddrs?: string[];
+  /**
+   * The responder's live STRAND-network multiaddrs for the provisioned strand — the
+   * initiator's only cross-party discovery seed, since the strand-addr RPC that resolves
+   * a sibling's strand addrs is membership-gated and answers own-party callers only.
+   *
+   * Disclosed on the same terms as `partyId`/`cadrePeerAddrs`: only after token +
+   * disclosure validation AND an approving provisioning outcome; a rejection carries
+   * none. Signaling-first, each entry ending in `/p2p/<responder strand transport
+   * peerId>` so the initiator can attribute it. OMITTED when the responder holds no live
+   * strand node for the provisioned strand (the responder-provisions path mints a strand
+   * that has not launched yet) — an absent or empty list is legal and means "no
+   * cross-party seed", never a protocol error.
+   */
+  strandAddrs?: string[];
+  /** The provisioned strand/db result (always present on approval). */
+  provisionResult?: FormationProvisionResult;
+}
+
+/**
+ * Normalize a strand-address list arriving from — or heading to — the wire: keep only
+ * parsable multiaddr strings, drop duplicates, and cap at {@link MAX_STRAND_ADDRS}.
+ * Order is preserved, so the responder's signaling-first ordering survives.
+ *
+ * Applied on BOTH sides on purpose. The responder bounds what it sends; the initiator
+ * bounds what it stores, because a peer is free to ignore the cap. A malformed entry is
+ * SKIPPED rather than fatal — these are runtime-discovered addresses (same posture as
+ * `extractCircuitRelayTargets` in `delegate-admission.ts`), and one bad entry must not
+ * cost a formation that has otherwise succeeded.
+ */
+export function sanitizeStrandAddrs(addrs: unknown): string[] {
+  if (!Array.isArray(addrs)) return [];
+  const out: string[] = [];
+  const seen = new Set<string>();
+  for (const entry of addrs) {
+    if (out.length >= MAX_STRAND_ADDRS) break;
+    if (typeof entry !== 'string' || entry.length === 0) {
+      log('strandAddrs: skipping non-string entry %o', entry);
+      continue;
+    }
+    if (seen.has(entry)) continue;
+    try {
+      multiaddr(entry);
+    } catch (error) {
+      // Deliberately not echoing the raw value at error level: it arrived from a remote
+      // peer. The log line below is the observability for a responder announcing junk.
+      log('strandAddrs: skipping unparsable entry: %o', error);
+      continue;
+    }
+    seen.add(entry);
+    out.push(entry);
+  }
+  return out;
+}
+
+// ── Stream framing helpers ───────────────────────────────────────────────────
+
+/**
+ * Reads length-prefixed JSON frames from a libp2p stream one frame at a time.
+ *
+ * Unlike seed delivery (which reads a stream to EOF then parses once), formation
+ * is request/response on a single live stream, so reading must stop at each frame
+ * boundary. The reader buffers across chunk boundaries and retains any overflow
+ * (a following frame coalesced into the same chunk) for the next read.
+ */
+class FrameReader {
+  private readonly iterator: AsyncIterator<Uint8Array>;
+  private readonly maxLength: number;
+  private buffer = new Uint8Array(0);
+
+  constructor(stream: AsyncIterable<Uint8Array>, maxLength = MAX_FORMATION_MSG_SIZE) {
+    this.iterator = stream[Symbol.asyncIterator]();
+    this.maxLength = maxLength;
+  }
+
+  private append(chunk: Uint8Array): void {
+    const next = new Uint8Array(this.buffer.length + chunk.length);
+    next.set(this.buffer, 0);
+    next.set(chunk, this.buffer.length);
+    this.buffer = next;
+  }
+
+  /** Pull chunks until the buffer holds ≥ `need` bytes; false if the stream ended first. */
+  private async fill(need: number): Promise<boolean> {
+    while (this.buffer.length < need) {
+      const { value, done } = await this.iterator.next();
+      if (done) return false;
+      const bytes = value instanceof Uint8Array
+        ? value
+        : (value as { subarray(): Uint8Array }).subarray();
+      this.append(bytes);
+    }
+    return true;
+  }
+
+  async read<T>(): Promise<T> {
+    if (!(await this.fill(4))) {
+      throw new Error('Formation stream closed before length prefix');
+    }
+    const length = new DataView(this.buffer.buffer, this.buffer.byteOffset, this.buffer.byteLength).getUint32(0, false);
+    if (length > this.maxLength) {
+      throw new Error(`Formation frame declares length ${length} exceeding max ${this.maxLength}`);
+    }
+    if (!(await this.fill(4 + length))) {
+      throw new Error(`Formation frame declares length ${length} but stream ended early`);
+    }
+    const json = new TextDecoder().decode(this.buffer.subarray(4, 4 + length));
+    this.buffer = this.buffer.subarray(4 + length);
+    return JSON.parse(json) as T;
+  }
+}
+
+// ── Result validation ────────────────────────────────────────────────────────
+
+/** Matches the placeholder cadre addrs (`cadre-a-1.local`, …) of the removed strand-proto transport. */
+const PLACEHOLDER_CADRE_ADDR = /^cadre-[ab]-\d+\.local$/;
+
+/**
+ * Structural check on a `responderCreates` result: the responder must have approved,
+ * disclosed a non-empty real identity + cadre (not the legacy `cadre-*.local`
+ * placeholders), and returned a strand vouched for by the responder party with a real id.
+ *
+ * `createdBy: 'responder'` means "the responder party vouches for / returns this strand."
+ * Under provision-then-record that strand was minted owner-signed earlier (not created
+ * in-session), but it is still the responder party returning its own strand, so the marker
+ * — and this structural validator — stay unchanged.
+ *
+ * The behavioral floor for the initiator: a responder that returns an arbitrary/empty
+ * `strandId` or omits its disclosed identity is rejected, not silently accepted.
+ */
+export function isValidResponderCreatesResult(response: FormationResultMessage): boolean {
+  if (!response.approved) return false;
+  if (!response.partyId) return false;
+  const addrs = response.cadrePeerAddrs;
+  if (!addrs || addrs.length === 0) return false;
+  if (addrs.some((a) => PLACEHOLDER_CADRE_ADDR.test(a))) return false;
+  const provision = response.provisionResult;
+  if (!provision?.strand?.strandId) return false;
+  if (provision.strand.createdBy !== 'responder') return false;
+  return true;
+}
+
+/**
+ * Ceiling for a resolved provisioning budget.
+ *
+ * The session budget also has to cover the wire step that PRECEDES provisioning (the
+ * responder's contact read, the initiator's dial-connect), so a whole step is held back
+ * rather than triggering only on a literal overrun.
+ *
+ * `reserveMs` is the EXTRA room only the responder holds back, so its own budget still
+ * lands strictly before the initiator's larger await-response budget even when both are
+ * clamped — without it, any configured budget at or above `sessionTimeoutMs -
+ * stepTimeoutMs` clamps both roles onto the same number and the responder's clean
+ * rejection races the initiator's timeout again. Capped at half the remaining room (like
+ * {@link splitProvisionBudget}) so a small session config still spends most of it working.
+ */
+function provisionCeilingMs(sessionTimeoutMs: number, stepTimeoutMs: number, reserveMs: number): number {
+  const roomMs = sessionTimeoutMs - stepTimeoutMs;
+  return Math.max(1, roomMs - Math.min(reserveMs, Math.floor(roomMs / 2)));
+}
+
+/**
+ * Resolve a caller-supplied provisioning budget: `0`/negative means "unset" (use
+ * `defaultMs`). If the result would let provisioning outlive the session — no result
+ * frame is ever sent, exactly the failure this budget exists to prevent — clamp it to
+ * {@link provisionCeilingMs} and log a warning.
+ */
+function resolveProvisionTimeoutMs(
+  configured: number | undefined,
+  defaultMs: number,
+  sessionTimeoutMs: number,
+  stepTimeoutMs: number,
+  role: string,
+  reserveMs = 0
+): number {
+  const requested = configured && configured > 0 ? configured : defaultMs;
+  const ceilingMs = provisionCeilingMs(sessionTimeoutMs, stepTimeoutMs, reserveMs);
+  if (requested >= ceilingMs) {
+    log(
+      '%s provisionTimeoutMs %dms leaves no room under sessionTimeoutMs %dms (step %dms, reserve %dms); clamping to %dms',
+      role, requested, sessionTimeoutMs, stepTimeoutMs, reserveMs, ceilingMs
+    );
+    return ceilingMs;
+  }
+  return requested;
+}
+
+/**
+ * Split a resolved provisioning budget into the WORK budget and the trailing settle grace.
+ *
+ * The grace is carved OUT of the budget (see {@link PROVISION_SETTLE_GRACE_MS}), and capped at
+ * half of it so a small configured budget still spends at least half its time doing work.
+ */
+function splitProvisionBudget(provisionTimeoutMs: number): { workMs: number; graceMs: number } {
+  const graceMs = Math.min(PROVISION_SETTLE_GRACE_MS, Math.floor(provisionTimeoutMs / 2));
+  return { workMs: provisionTimeoutMs - graceMs, graceMs };
+}
+
+// ── Responder (listener) ─────────────────────────────────────────────────────
+
+export interface FormationListenerOptions {
+  /** Validate the invitation token; returns whether it is valid. */
+  validateToken(token: string): Promise<{ valid: boolean }>;
+  /** Validate the initiator's disclosure with the REAL token + disclosure. */
+  validateDisclosure(token: string, disclosure: StrandFormationDisclosure): Promise<boolean>;
+  /**
+   * Provision (or, for provision-then-record, resolve + record consent against) the
+   * strand (`responderCreates`) for the given initiator. The WHOLE contact message is
+   * threaded in: the hook needs the REAL token to map the bound host strand, and the
+   * joiner's `peerKey`/`usageStampId`/`peerSignature` to write the consent columns.
+   *
+   * Returns a {@link ResponderProvisionOutcome}: the hook may REJECT post-validation
+   * (e.g. an unconverged host strand) and the listener turns that into a clean,
+   * non-disclosing `approved: false` reply rather than dropping the result frame.
+   *
+   * `signal` is aborted when the listener's work budget expires: a hook that observes it
+   * BEFORE writing abandons the redemption and leaves the invite unspent. Optional so
+   * signal-unaware hooks (tests, mocks) stay assignable.
+   */
+  provisionStrand(contact: FormationContactMessage, signal?: AbortSignal): Promise<ResponderProvisionOutcome>;
+  /** Responder identity, disclosed only AFTER token + disclosure validation passes. */
+  getResponderIdentity(): { partyId: string; cadrePeerAddrs: string[] };
+  /**
+   * This node's live STRAND-network multiaddrs for a provisioned strand, read on the
+   * approval path only (same disclosure timing as {@link getResponderIdentity}) and
+   * carried back as {@link FormationResultMessage.strandAddrs}.
+   *
+   * Optional, and an empty answer is legal: the strand may not be running here yet (the
+   * responder-provisions path mints it during this very session). The listener cannot
+   * reach the strand runtime itself, so this is the seam `CadreNode` wires to its own
+   * per-strand address lookup.
+   */
+  resolveStrandAddrs?(strandId: string): string[];
+  sessionTimeoutMs?: number;
+  stepTimeoutMs?: number;
+  /**
+   * Budget for the `provisionStrand` hook call — distinct from `stepTimeoutMs` because
+   * provisioning is real work, not a bare wire read/write. Default 12 s; `0`/unset uses
+   * the default; a value that would outlive the session is clamped ({@link
+   * provisionCeilingMs}). See {@link DEFAULT_PROVISION_TIMEOUT_MS} for the full ordering
+   * rationale.
+   */
+  provisionTimeoutMs?: number;
+  maxConcurrentSessions?: number;
+}
+
+/**
+ * Responder side of the native formation protocol. Registers a libp2p handler and
+ * drives each inbound stream through token + disclosure validation, provisioning,
+ * and result delivery — enforcing the cadre-disclosure timing rule (responder cadre
+ * is revealed only after validation; rejections disclose nothing).
+ */
+export class FormationListener {
+  private readonly options: FormationListenerOptions;
+  private readonly sessionTimeoutMs: number;
+  private readonly stepTimeoutMs: number;
+  private readonly provisionWorkMs: number;
+  private readonly provisionGraceMs: number;
+  private readonly maxConcurrentSessions: number;
+  private readonly registered = new Set<Libp2p>();
+  private activeSessions = 0;
+  private sessionCounter = 0;
+
+  constructor(options: FormationListenerOptions) {
+    this.options = options;
+    this.sessionTimeoutMs = options.sessionTimeoutMs ?? DEFAULT_SESSION_TIMEOUT_MS;
+    this.stepTimeoutMs = options.stepTimeoutMs ?? DEFAULT_STEP_TIMEOUT_MS;
+    const split = splitProvisionBudget(resolveProvisionTimeoutMs(
+      options.provisionTimeoutMs, DEFAULT_PROVISION_TIMEOUT_MS,
+      this.sessionTimeoutMs, this.stepTimeoutMs, 'FormationListener',
+      PROVISION_RESPONSE_TRAVEL_MARGIN_MS
+    ));
+    this.provisionWorkMs = split.workMs;
+    this.provisionGraceMs = split.graceMs;
+    this.maxConcurrentSessions = options.maxConcurrentSessions ?? DEFAULT_MAX_CONCURRENT_SESSIONS;
+  }
+
+  /** Number of in-flight inbound sessions. */
+  get activeCount(): number {
+    return this.activeSessions;
+  }
+
+  register(node: Libp2p, protocolId: string = FORMATION_PROTOCOL): void {
+    if (this.registered.has(node)) {
+      log('node already registered');
+      return;
+    }
+    void node.handle(protocolId, async (rawStream: unknown, _connection: unknown) => {
+      await this.handleStream(rawStream as ControlStream);
+    });
+    this.registered.add(node);
+    log('formation listener registered (%s)', protocolId);
+  }
+
+  unregister(node: Libp2p, protocolId: string = FORMATION_PROTOCOL): void {
+    if (!this.registered.has(node)) return;
+    void node.unhandle(protocolId);
+    this.registered.delete(node);
+    log('formation listener unregistered (%s)', protocolId);
+  }
+
+  private async handleStream(stream: ControlStream): Promise<void> {
+    if (this.activeSessions >= this.maxConcurrentSessions) {
+      const rejection: FormationResultMessage = { approved: false, reason: 'Too many concurrent formation sessions' };
+      try { writeFrame(stream, rejection); } catch { /* best effort */ }
+      try { await stream.close(); } catch { /* ignore */ }
+      return;
+    }
+    const id = ++this.sessionCounter;
+    this.activeSessions++;
+    try {
+      await withTimeout(this.sessionTimeoutMs, `Formation session#${id}`, () => this.runSession(id, stream));
+    } catch (err) {
+      log('formation session #%d failed: %o', id, err);
+    } finally {
+      this.activeSessions--;
+      try { await stream.close(); } catch { /* ignore */ }
+    }
+  }
+
+  /**
+   * Run the provisioning hook under its WORK budget — the configured provisioning budget minus
+   * the trailing settle grace ({@link splitProvisionBudget}, {@link PROVISION_SETTLE_GRACE_MS}).
+   *
+   * When the work budget expires the hook's signal is aborted, then
+   * {@link settleWithinGrace} waits out the grace for the aborted work to settle anyway.
+   * Returns `undefined` for "timed out" — which the caller turns into a retryable rejection
+   * frame — and rethrows every other failure so it reaches the internal-error path.
+   *
+   * Deliberately not `withDeadline`: that exposes neither the timed-out flag nor the signal
+   * outside its `op`, and both are needed here.
+   */
+  private async provision(id: number, contact: FormationContactMessage): Promise<ResponderProvisionOutcome | undefined> {
+    const controller = new AbortController();
+    let timedOut = false;
+    const pending = this.options.provisionStrand(contact, controller.signal);
+    try {
+      return await withTimeout(
+        this.provisionWorkMs,
+        `Formation provisioning#${id}`,
+        () => pending,
+        () => { timedOut = true; controller.abort(); }
+      );
+    } catch (err) {
+      if (!timedOut) throw err;
+      log(
+        'formation session #%d provisioning work budget expired after %dms; aborted, settling up to %dms',
+        id, this.provisionWorkMs, this.provisionGraceMs
+      );
+      return await this.settleWithinGrace(id, pending);
+    }
+  }
+
+  /**
+   * Wait out the settle grace for an ABORTED provisioning, adopting a late success.
+   *
+   * A hook that observed the abort before issuing its `FormationUsage` insert rejects (with
+   * `FormationAbortedError`) and leaves the invite unspent → reported as a timeout. A hook
+   * that had already written cannot be un-written (the table is append-only), so if it lands
+   * inside the grace its outcome is ADOPTED — the joiner is told the truth (approved, or the
+   * hook's own rejection reason) rather than "timed out" over a spent invite.
+   *
+   * NOTE: one window stays open — a work budget that expires after the `FormationUsage`
+   * insert was issued AND a commit that outlasts the grace. The joiner is told 'timed out'
+   * while its one-time invite is in fact spent, and since every retry mints a fresh keypair
+   * no recovery path can match it. Nothing can un-spend an append-only row, so the late-settle
+   * logging below is the observability for it; widen the grace only if it is seen in the wild.
+   */
+  private async settleWithinGrace(
+    id: number,
+    pending: Promise<ResponderProvisionOutcome>
+  ): Promise<ResponderProvisionOutcome | undefined> {
+    let stillPending = false;
+    try {
+      const outcome = await withTimeout(
+        this.provisionGraceMs,
+        `Formation settle#${id}`,
+        () => pending,
+        () => { stillPending = true; }
+      );
+      log('formation session #%d provisioning settled within its grace: approved=%s — adopting', id, outcome.approved);
+      return outcome;
+    } catch (err) {
+      if (!stillPending) {
+        log('formation session #%d provisioning failed after its abort: %o', id, err);
+        return undefined;
+      }
+      log('formation session #%d provisioning still pending after its %dms grace', id, this.provisionGraceMs);
+      // Log how it eventually settled so a late failure is not swallowed by the two
+      // abandoned `withTimeout` calls.
+      void pending.then(
+        (late) => log('formation session #%d provisioning settled after its grace: approved=%s', id, late.approved),
+        (err2) => log('formation session #%d provisioning failed after its grace: %o', id, err2)
+      );
+      return undefined;
+    }
+  }
+
+  /**
+   * Joiner-consent pre-check — the legibility twin of the schema's
+   * `FormationUsage.PeerConsented` CHECK, exactly as `verifyFormationApproval` is for the
+   * approver's vouch. The database constraint stays the authority; this turns a bad consent
+   * into a clean 'Invalid joiner consent' rejection instead of an opaque CHECK failure at
+   * commit. Runs BEFORE token/disclosure validation: it is the cheapest check (pure local
+   * crypto, no DB read) and a rejection discloses nothing.
+   *
+   * Also pins `partyId` to `peerKey`: an ed25519 peer id is an identity multihash of the
+   * very key, and this is the ONLY layer that can check the match — SQL cannot unwrap a
+   * multihash, and `partyId` itself is never stored.
+   */
+  private isJoinerConsentValid(id: number, contact: FormationContactMessage): boolean {
+    try {
+      requireEd25519PublicKeyB64(contact.peerKey, 'joiner peer key');
+    } catch {
+      // Deliberately not rethrown or echoed: the helper's error message embeds the whole
+      // rejected value, and this one arrived from a remote peer (the cap-the-echo case its
+      // own NOTE warns about). Log without the raw value instead.
+      log('formation session #%d: joiner peer key is not a base64url ed25519 public key', id);
+      return false;
+    }
+    const embedded = ed25519PublicKeyB64FromPeerId(contact.partyId);
+    if (!embedded || embedded !== contact.peerKey) {
+      log('formation session #%d: partyId does not embed the presented peer key', id);
+      return false;
+    }
+    const consentOk = verifyFormationConsent({
+      token: contact.token,
+      usageStampId: contact.usageStampId,
+      peerKey: contact.peerKey,
+      disclosure: canonicalJson(contact.disclosure),
+      peerSig: contact.peerSignature
+    });
+    if (!consentOk) {
+      log('formation session #%d: joiner consent signature failed verification', id);
+      return false;
+    }
+    return true;
+  }
+
+  /**
+   * The strand addresses to disclose for an APPROVED provisioning, sanitized and capped.
+   *
+   * Never throws: a hook that fails (a strand runtime torn down mid-session) costs the
+   * joiner its cross-party seed, not the formation it has already earned — the seed is
+   * an optimization, the consent row is the commitment.
+   */
+  private strandAddrsFor(id: number, strandId: string): string[] {
+    if (!this.options.resolveStrandAddrs) return [];
+    try {
+      return sanitizeStrandAddrs(this.options.resolveStrandAddrs(strandId));
+    } catch (err) {
+      log('formation session #%d: strand-addr lookup for %s failed (continuing): %o', id, strandId, err);
+      return [];
+    }
+  }
+
+  private async runSession(id: number, stream: ControlStream): Promise<void> {
+    // Track whether ANY frame has been written so the catch below can convert an
+    // unexpected internal error into a non-disclosing rejection ONLY when nothing has
+    // gone out yet. This closes the "stream closed with no result frame" class of bug.
+    let wroteFrame = false;
+    const send = (msg: FormationResultMessage): void => {
+      writeFrame(stream, msg);
+      wroteFrame = true;
+    };
+
+    try {
+      const reader = new FrameReader(stream);
+      const contact = await withTimeout(this.stepTimeoutMs, `Formation await-contact#${id}`, () => reader.read<FormationContactMessage>());
+      log('formation session #%d contact: token=%s party=%s', id, contact.token, contact.partyId);
+
+      if (!this.isJoinerConsentValid(id, contact)) {
+        send({ approved: false, reason: 'Invalid joiner consent' });
+        return;
+      }
+
+      const tokenResult = await this.options.validateToken(contact.token);
+      if (!tokenResult.valid) {
+        send({ approved: false, reason: INVALID_TOKEN_REASON });
+        return;
+      }
+
+      const disclosureOk = await this.options.validateDisclosure(contact.token, contact.disclosure);
+      if (!disclosureOk) {
+        send({ approved: false, reason: 'Invalid disclosure' });
+        return;
+      }
+
+      const outcome = await this.provision(id, contact);
+      if (!outcome) {
+        // Reported as its own retryable reason rather than falling through to the generic
+        // catch below, which would report the misleading 'Internal formation error'.
+        send({ approved: false, reason: 'Formation provisioning timed out' });
+        return;
+      }
+      if (!outcome.approved) {
+        // A post-validation rejection still discloses NEITHER identity NOR cadre,
+        // exactly like the token/disclosure rejections above.
+        send({ approved: false, reason: outcome.reason });
+        return;
+      }
+      // Validation + provisioning passed → safe to disclose responder identity/cadre.
+      const identity = this.options.getResponderIdentity();
+      const strandAddrs = this.strandAddrsFor(id, outcome.result.strand.strandId);
+      send({
+        approved: true,
+        partyId: identity.partyId,
+        cadrePeerAddrs: identity.cadrePeerAddrs,
+        // Omitted rather than sent empty: an absent field and an empty list mean the same
+        // thing to the initiator, and omitting keeps every rejection-parity assertion a
+        // plain `toBeUndefined()`.
+        ...(strandAddrs.length > 0 ? { strandAddrs } : {}),
+        provisionResult: outcome.result
+      });
+    } catch (err) {
+      // Defense-in-depth: if an unexpected internal error escapes BEFORE any frame was
+      // written (e.g. a future provisionStrand-hook bug), convert it into a non-disclosing
+      // protocol rejection so the initiator sees a clean error instead of a read-error/
+      // timeout. If a frame was already sent, leave the stream as-is and let handleStream
+      // log + close. Re-throw either way so the failure is still recorded — this is a
+      // deliberate, logged conversion, not silent.
+      if (!wroteFrame) {
+        const internalError: FormationResultMessage = { approved: false, reason: 'Internal formation error' };
+        try { writeFrame(stream, internalError); } catch { /* stream already broken */ }
+      }
+      throw err;
+    }
+  }
+}
+
+// ── Initiator (dialer) ───────────────────────────────────────────────────────
+
+export interface FormationDialOptions {
+  /** The contact message to send (real token, partyId, disclosure, initiator cadre). */
+  contact: FormationContactMessage;
+  /** Responder multiaddrs to dial. */
+  responderAddrs: string[];
+  /** Validate the responder's result; a false return aborts the formation. */
+  validateResponse(response: FormationResultMessage): Promise<boolean>;
+  sessionTimeoutMs?: number;
+  stepTimeoutMs?: number;
+  /**
+   * Budget for the `await-response` read only (the responder's provisioning + reply travel
+   * time) — `dial-connect` keeps using `stepTimeoutMs`. Default 15 s; `0`/unset uses the
+   * default. See {@link DEFAULT_PROVISION_TIMEOUT_MS} for the full ordering rationale.
+   */
+  provisionTimeoutMs?: number;
+  protocolId?: string;
+}
+
+/**
+ * What one successful formation dial yields the initiator.
+ *
+ * The provisioned strand travels in {@link FormationProvisionResult}, but the responder's
+ * strand-network addresses are NOT part of that structure — they are a disclosure of the
+ * responder's live runtime, alongside `partyId`/`cadrePeerAddrs`, not a property of the
+ * strand row. Returning them beside the provision result keeps that separation while
+ * still handing the caller the one thing it needs to seed a cross-party mesh.
+ *
+ * `strandAddrs` is always an array and is frequently EMPTY — see
+ * {@link FormationResultMessage.strandAddrs}.
+ */
+export interface FormationDialResult {
+  provision: FormationProvisionResult;
+  /** Sanitized + capped ({@link sanitizeStrandAddrs}); `[]` when the responder sent none. */
+  strandAddrs: string[];
+}
+
+/**
+ * Initiator side of the native formation protocol. Dials the responder, sends the
+ * contact (carrying the real disclosure/token/cadre), validates the responder's
+ * result, and returns the strand the responder provisioned plus the responder's
+ * strand-network addresses (see {@link FormationDialResult}).
+ */
+export async function dialFormation(node: Libp2p, options: FormationDialOptions): Promise<FormationDialResult> {
+  if (options.responderAddrs.length === 0) {
+    throw new Error('No responder addresses available for formation');
+  }
+  const protocolId = options.protocolId ?? FORMATION_PROTOCOL;
+  const sessionTimeoutMs = options.sessionTimeoutMs ?? DEFAULT_SESSION_TIMEOUT_MS;
+  const stepTimeoutMs = options.stepTimeoutMs ?? DEFAULT_STEP_TIMEOUT_MS;
+  const provisionTimeoutMs = resolveProvisionTimeoutMs(
+    options.provisionTimeoutMs, DEFAULT_INITIATOR_PROVISION_TIMEOUT_MS,
+    sessionTimeoutMs, stepTimeoutMs, 'dialFormation'
+  );
+  const addr = multiaddr(options.responderAddrs[0]);
+
+  return withTimeout(sessionTimeoutMs, 'Formation dial', async () => {
+    const rawStream = await withTimeout(stepTimeoutMs, 'Formation dial-connect', async () => node.dialProtocol(addr, protocolId));
+    const stream = rawStream as unknown as ControlStream;
+    try {
+      const reader = new FrameReader(stream);
+      writeFrame(stream, options.contact);
+
+      const response = await withTimeout(provisionTimeoutMs, 'Formation await-response', () => reader.read<FormationResultMessage>());
+      if (!response.approved) {
+        throw new Error(`Formation rejected: ${response.reason ?? 'no reason provided'}`);
+      }
+      const ok = await options.validateResponse(response);
+      if (!ok) throw new Error('Responder result failed validation');
+
+      if (!response.provisionResult) throw new Error('Missing provision result for responderCreates mode');
+      // Sanitized HERE — the wire boundary — so nothing above this call ever handles an
+      // unvalidated, unbounded address list from a remote peer.
+      return { provision: response.provisionResult, strandAddrs: sanitizeStrandAddrs(response.strandAddrs) };
+    } finally {
+      try { await stream.close(); } catch { /* ignore */ }
+    }
+  });
+}