✔ agents: a held role with nothing on the bus is idle; a role session gets the claim rows unchanged (109.612256ms)
✔ agents: working counts what the holder received and has not read; routed alone is not working (205.729691ms)
✔ agents: input needed is the role's open decisions routed to the human, gated first, then oldest (335.682307ms)
✔ agents: working wins over waiting, and Input needed still shows (165.623333ms)
✔ agents: a new holder doesn't inherit what the last one left unread; decisions stay with the role (179.405427ms)
✔ agents: state reads only its own business, even for a run name both use (197.356673ms)
✔ agents: a message the holder released unread is stranded; the next holder gets nothing and cannot read it (200.072357ms)
✔ agents: a message the holder read is not stranded; one the current holder has not read is not listed (195.270516ms)
✔ agents: stranded is oldest first and capped at 50 for a role (834.236942ms)
✔ agents: stranded is per role and per business (298.202247ms)
✔ agents: a message routed to a held role and not yet received is not stranded (117.199279ms)
✔ launch identity is stamped, payload identity is refused and stale holder cannot send (151.415601ms)
✔ decision classes route from policy; gated resolution is human-only, choice and target must match (260.040866ms)
✔ claim exclusion, holder release, gated revoke and rerouting to a new holder are atomic (280.404818ms)
✔ launch events require a human CLI capability; generic emit cannot forge authority events (165.14355ms)
✔ within-role decisions close atomically and invalid options or blocking omissions refuse (171.151677ms)
✖ a recommendation that is not exactly one option key refuses invalid-recommendation after the options (194.212713ms)
✔ observer capabilities read human inbox but cannot mutate or forge launch identity (177.751711ms)
✔ task action subjects and linked decision trail are complete and ordered (185.19045ms)
✔ launch binding is durable and reconnecting requires the identical trusted record (125.442726ms)
✔ business isolation includes inherited object names and cross-business message references (181.652793ms)
✔ authority never transfers between action, run, target, unresolved or replaced role holder (270.014792ms)
✔ task projection uses schema current view, skipping earlier and equal-start polls (159.030444ms)
✔ revocation permanently bars the old run from reclaiming first, including after broker restart (198.301297ms)
✔ empty message references refuse before storage; refusal-evidence failure stays a typed error (135.992011ms)
✔ both arbiters require human resolution when their cross-role route is themselves (204.010532ms)
✔ the shipped coder role raises task.reassign as gated to the human; task.scope.change stays cross-role (137.876312ms)
✔ S1 adapter takes resolved limits and refs, rejects mismatched instance, never mutates input (2.013294ms)
✔ credentials gives human and reader sessions service, account, role, date and state, and nothing else (92.458403ms)
✔ credentials refuses a role session and any argument (170.011951ms)
✔ credentials reads the status on each request, and a broker without one lists none (234.580379ms)
✔ the broker process answers credentials without a token, path or variable in its output, log or store (183.505491ms)
✔ only validated broker references load; returned data and exceptions cannot expose a known token (3.498804ms)
✔ bad file modes, symlinks, repository/data paths, malformed tokens and missing dates refuse (3.049743ms)
✔ expiry refuses use and env references never become client data (0.741351ms)
✔ S1 parsed service refs work, service mismatch refuses, Gitea rotation due is a warning state (1.361679ms)
✔ opaque tokens shorter than 16 characters refuse before use (0.279186ms)
✔ roles lists open decisions routed to a role, and never the human inbox (216.10926ms)
✔ recent lists closed decisions of every route, newest close first, with how each closed (297.026269ms)
✔ recent gives the closing event as closed when the decision was seen first (292.00375ms)
✔ recent stops at RECENT rows (772.513987ms)
✔ every row is what trail already gives the same session for that id (257.386489ms)
✔ a role session can't list decisions, so it can't read another role's open decisions (376.267367ms)
✔ decisions is a read: one business, strict arguments, nothing written (368.915448ms)
✔ endLaunch writes session.ended, releases the run claim and kills its capabilities (123.401715ms)
✔ endLaunch refuses an unknown run, a second end and a rebind of the ended run (138.170725ms)
✔ a restarted broker refuses to rebind an ended run; a refused rebind leaves the run unbound (162.222456ms)
✔ endLaunch leaves a claim another run took alone (189.261156ms)
✔ refuse records action.refused against the caller with the code only (127.268756ms)
✔ launches off refuses role.launch with launch-revoked until launches on (189.47551ms)
✔ broker process: launch ops authorize role.launch, record refusals and end runs (248.53352ms)
✔ human proof binds CLI entry, process start and nonce; agents and incomplete ancestry refuse (1.897402ms)
✔ process reader gets own kernel identity without exposing environment values (0.549139ms)
✔ EACCES ancestor environments skip only markers; commands and registered launches still refuse (0.740532ms)
✔ real pid 1 remains inspectable when its environment is protected (0.265327ms)
✔ within-role sends cite an open gated launch decision without spending it or naming it in grants (186.278376ms)
✔ missing and foreign-business citations refuse and roll back message and grant (187.548768ms)
✔ cross-role sends still need a matching resolved decision and consume it once (244.464601ms)
✔ messages pages all human messages newest first without writes; read state is history (2479.998065ms)
✔ messages refuses role callers, invalid paging and cursors outside this human feed (236.098076ms)
✔ views messages passes only its paging arguments through (0.373594ms)
✔ messages isolates businesses even with a valid foreign cursor (141.974691ms)
✔ broker process binds trusted launches, offers reader capabilities, refuses human mutation, closes cleanly (183.761662ms)
✔ startup token refusal returns safe code without value or partial listening broker (40.657441ms)
✔ loaded fixture token is absent from socket replies and SQLite, including refusal evidence (174.713621ms)
✔ killed broker leaves an explicit stale lock; another process cannot silently reclaim it (179.54616ms)
✔ trusted host registers later launches; socket clients never have a registration verb (188.401712ms)
✔ runtime excludes declared project roots even when host supplies no repoRoots (34.343252ms)
✔ a refused launch binding leaves the broker and existing capabilities alive; bad protocol stops it (162.639976ms)
✔ a reply is listed for human and reader sessions, and its delivery rows are unchanged (228.910859ms)
✔ the reply still reaches the human message.receive (mosaic talk) after the Console showed it (203.707148ms)
✔ a message with no reply gives an empty list (172.309703ms)
✔ replies come oldest first and stop at REPLIES (737.056872ms)
✔ an id that names no message in the business refuses message-not-found, and the refusal is recorded (284.531638ms)
✔ strict arguments; a role session gets human-required (344.991263ms)
✔ views passes replies through as one call with the id (0.289911ms)
✔ requirements returns the committed list to a holding role session, and no one else (216.892937ms)
✔ the PM's prd.approve with no task, when no id fits, reaches the human's Inbox (164.667883ms)
✔ the PM's non-blocking role.revoke with no task reaches the human's Inbox (184.614234ms)
✔ a push is recorded as the action.allowed for git.push.working on the task, and only the host records one (236.022482ms)
✔ review.request names the latest pushed commit and sends the reviewer a REVIEW-REQUEST (254.8966ms)
✔ review.request says which run holds the reviewer role, or null, and a null request waits for a claim (291.229963ms)
✔ review.request refuses a commit that isn't the latest push, a non-reviewer and a role without the action (356.213633ms)
✔ reviewFetch answers only the addressed reviewer, for the request it cites (247.562721ms)
✔ review.verdict answers one request once, to the role that asked (289.423474ms)
✔ task.close on a pushed task needs an approve verdict for the latest push (342.346251ms)
✔ a refusal the host decided is filed on the task it names, and only a task ref counts (244.490684ms)
✔ v3b prototype refusals, views and append-only mutations (1041.504745ms)
✔ gated approval authorizes once, survives store reopen, and fresh approval works (263.375992ms)
✔ another run cannot consume an approval; a failed check leaves it usable (225.958941ms)
✔ two scheduled callers have exactly one grant and one consumed refusal (152.990276ms)
✔ failed commit rolls consumption back; cross-role consumes and within-role stays reusable (323.819589ms)
✔ class drift gated to cross-role refuses before consumption (234.072704ms)
✔ class drift cross-role to gated refuses before consumption (246.399908ms)
✔ class drift gated to within-role refuses before consumption (229.136022ms)
✔ class drift cross-role to within-role refuses before consumption (228.566197ms)
✔ class drift within-role to gated refuses before consumption (189.512956ms)
✔ class drift within-role to cross-role refuses before consumption (203.921249ms)
✔ message.send consumes approval and prevents a later send or authorize (198.458163ms)
✔ role.revoke consumes approval and prevents a later revoke or authorize (229.922313ms)
✔ a business that never switched launches reads allowed with no time and no event (134.06685ms)
✔ launches revoked and then restored read allowed, with the restore time and event id (150.124297ms)
✔ a task closed on the tracker leaves the open list, and total counts the open ones (148.079986ms)
✔ a tombstone is not open (154.640613ms)
✔ the open list stops at the cap and says so; total still counts every open task (136.75624ms)
✔ a role and a reader session read state; arguments, and an unbound cap, refuse (142.899413ms)
✔ a human session reads the same state, with the same cap (146.49048ms)
✔ reading state writes nothing: every table keeps its row count (159.434617ms)
✔ creates private WAL store and excludes a second writer until explicit close (136.747459ms)
✔ rollback is atomic and schema metadata is checked against trusted DDL, not just itself (164.671829ms)
✔ existing empty database and symlink runtime directory refuse, never initialize over damage (194.063523ms)
✔ crash during a transaction recovers no partial event after explicit fixture-only lock removal (158.372108ms)
✔ writer refuses mixed at/read_at forms atomically, even through trusted SQL helpers (116.259499ms)
✔ async transactions refuse before invoking their function (98.667072ms)
✔ the session default fits under the reply cap and pages to the end with next (286.978281ms)
✔ the short form: no description, assignees by role, every listed key (239.626845ms)
✔ each filter, alone and together, and total counts every match (348.598561ms)
✔ task_ref reads one task in full, as the whole board has it (290.57225ms)
✔ no arguments is the read the Console and mosaic make, unchanged (246.635205ms)
✔ an unknown argument or a bad value refuses invalid-request (915.18312ms)
✔ tasks rows carry priority and requirement for every session that reads tasks (309.564555ms)
✔ the read model rebuilds from the ledger: a reopened bus and each trail give the same values (338.457355ms)
✔ an empty business and one with snapshots but no task.created read cleanly (117.098584ms)
✔ recordTask keeps sync reads and a role write apart (189.661832ms)
✔ read_at must be one canonical UTC format, so the projection compares strings safely (104.330886ms)
✔ a bad entry refuses the whole record (105.532405ms)
✔ taskView reads the projection for one business (147.516746ms)
✔ requestTask hands only a holder and a task verb to the handler, and records refusals (230.697766ms)
✔ the server sends task verbs to the adapter with its own timeout; other verbs stay synchronous (409.937823ms)
✔ without an adapter the server refuses every task verb (209.697333ms)
✔ the runtime refuses an invalid adapter and closes a valid one (235.423858ms)
✔ the process loads the S3 adapter from plain-data trackers (250.132971ms)
✔ the tester binds only when the business sets bus.tester to exactly true (760.220923ms)
✔ a business may not name a role tester, so the tester address is never a role (228.68382ms)
✔ every verb outside the tester list refuses tester-refused and is recorded against the tester (535.265045ms)
✔ a tester message is from tester, starts [mos test], and writes no human.input or authority (202.226729ms)
✔ a role answers a tester only in reply to a tester message, and the answer never reaches the human (231.321766ms)
✔ the tester reads the human views a reader reads (112.505119ms)
✔ tester proof binds its own CLI, start time and nonce, refuses a managed run and allows an agent (1.451719ms)
✔ neither CLI's nonce proves the other: each proof is bound to its own CLI and nonce variable (0.434562ms)
✔ testerNonce reads only the tester's variable and gives null on any read error (0.331239ms)
✔ an envelope carries exactly one of cap, human and tester; no tester authenticator refuses (56.062063ms)
✔ tester-cli.mjs binds through the runtime when the switch is on, and refuses off or under a run marker (799.393181ms)
✔ socket capability stamps launch identity; shared views use wire, no SQL client (163.61538ms)
✔ two wire claims serialize; a lost reply never automatically retries (168.326822ms)
✔ malformed, oversized and identity-forging envelopes refuse without echoing input (104.446232ms)
✔ client preserves UTF-8 when a response divides a multibyte character (12.747531ms)
✔ committed mutation followed by dropped reply reports unknown and is never retried (130.674107ms)
ℹ tests 148
ℹ suites 0
ℹ pass 147
ℹ fail 1
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 3972.299816

✖ failing tests:

test at tests/broker.test.mjs:158:1
✖ a recommendation that is not exactly one option key refuses invalid-recommendation after the options (194.212713ms)
  AssertionError [ERR_ASSERTION]: The input did not match the regular expression /invalid-options/. Input:
  
  'Error: invalid-recommendation'
  
      at TestContext.<anonymous> (file:///home/jwoltje/darkwing-scratch/r112/wt/packages/bus/tests/broker.test.mjs:174:10)
      at Test.runInAsyncScope (node:async_hooks:226:14)
      at Test.run (node:internal/test_runner/test:1402:25)
      at Test.processPendingSubtests (node:internal/test_runner/test:974:18)
      at Test.postRun (node:internal/test_runner/test:1542:19)
      at Test.run (node:internal/test_runner/test:1467:12)
      at async Test.processPendingSubtests (node:internal/test_runner/test:974:7) {
    generatedMessage: true,
    code: 'ERR_ASSERTION',
    actual: Error: invalid-recommendation
        at fail (file:///home/jwoltje/darkwing-scratch/r112/wt/packages/bus/src/broker.mjs:76:9)
        at #dispatch (file:///home/jwoltje/darkwing-scratch/r112/wt/packages/bus/src/broker.mjs:784:50)
        at file:///home/jwoltje/darkwing-scratch/r112/wt/packages/bus/src/broker.mjs:675:38
        at Store.transaction (file:///home/jwoltje/darkwing-scratch/r112/wt/packages/bus/src/store.mjs:156:22)
        at Broker.request (file:///home/jwoltje/darkwing-scratch/r112/wt/packages/bus/src/broker.mjs:669:26)
        at call (file:///home/jwoltje/darkwing-scratch/r112/wt/packages/bus/tests/broker.test.mjs:45:45)
        at raise (file:///home/jwoltje/darkwing-scratch/r112/wt/packages/bus/tests/broker.test.mjs:47:3)
        at file:///home/jwoltje/darkwing-scratch/r112/wt/packages/bus/tests/broker.test.mjs:174:23
        at getActual (node:assert:580:5)
        at strict.throws (node:assert:728:24),
    expected: /invalid-options/,
    operator: 'throws',
    diff: 'simple'
  }