+ WjrfRt^RIt^RIt^RIt^RIt^RIHtHt^RIH t ^RI H t H t H t HtHt] !]4P"P"P%4t]P(P+^]!]44^RIHtHtRRRlltRtR R ltR R ltR RltRRltRRlt RRlt!RRlt"RRlt#RRlt$RRlt%RR]PL3RR ]PN3/t(RR"R#llt)R$R%lt*R&R'lt+RR(R)llt,R*R+lt-R,R-lt.R.R/lt/R0t0RR1R2llt1R3]PdR4]PLR5]PfR6]Ph/t5R7R8lt6R9t7R:t8R;R<lt9R=R>lt:R?R!/R@RAllt;R?R!/RBRClltRGRHlt?RIRJlt@RKRLltARtBRMRNltCRORPltDRQRRltERtFRSRTltGRtHRURVltIRWtJRXtKRYRZltLR[R\ltMR]R^ltNR_R`ltORaRbRcRdReRfRgRhRiRjRkRlRmRnRoRpRqRrRsRtRuRvRwRxRyRz/ tP]PRi,]PR{&]PRo,]PR|&]PRy,;]PR}&]PR~&RtQR#)zCron subcommand for hermes CLI.N)timezoneUTCPath)AnyDictIterableListOptional)ColorscolorcV^8dQhR\\\,,R\\\,,/#)skillsreturn)r rstrlist)formats"CC:\Users\johnn\AppData\Local\hermes\hermes-agent\hermes_cli\cron.py __annotate__rs0  (3-1H T\]abe]fTg cVfVfR#.pVe \V4MV.FIp\T;'gR4P4pV'gK0WB9gK8VPV4KK V#)zDDeduped, stripped skill names; None when neither argument was given.N)rrstripappend) single_skillr normalizeditemtexts&& r_normalize_skillsrsg ~,.J & 2V F4::2$$& 4D*   d #G rc F^RIHp\P!V!R/VB4#))cronjob)tools.cronjob_toolsr"jsonloads)kwargs cronjob_tools, r _cron_apir)s; ::l,V, --rc$V^8dQhR\/#rrr)rs"rrr"sCrcj^RIHpV!4P;'gR# \dR#i;i)uXResolved cron scheduler provider name ('builtin', 'chronos', …); 'builtin' on failure.resolve_cron_schedulerbuiltin)cron.scheduler_providerr/name Exceptionr.s r_active_cron_provider_namer4"s3B%',,99 9 s## 22c:V^8dQhR\\,/#r+)r bool)rs"rrr+s8D>rc\4R8wdR#\P!\4;_uu_4^RIHpV!4'd RRR4R#RRR4^RIHpHpV!4'dR#^RI H pH p\V!44'gR#V!4;'gV!4# +'giL];i \dR#i;i)uTri-state readiness of a scheduler serving THIS home (None = probe failed). A gateway launched from this home proves itself by process liveness. Any other scheduler that ticks this home — the host multiplexer for a served profile, or the in-process ticker inside ``hermes serve`` / the Desktop backend (#121881) — proves itself only by the fresh heartbeat it writes into this home's store. External providers use their own machinery and are exempt. r0T)is_gateway_runtime_lock_activeNfind_gateway_pids+named_profile_served_by_running_multiplexerget_ticker_heartbeat_ageticker_heartbeat_writer_aliveF) r4 contextlibsuppressr3gateway.statusr8hermes_cli.gatewayr:r; cron.jobsr=r>_ticker_age_is_fresh)r8r:r;r=r>s r_builtin_gateway_livenessrE+s % '9 4  + + E-//, +/, f   U#$<$>?? ;<__@]@__, + sEB="B=B* B=B=4B= B="B=* B: 5B== C  C cV^8dQhRR/#rrNr#)rs"rrrLs G Gd Grc\4RJdR#^RIHp\\ RV!4 R2\ P 44\\ R\ P44R#)zaWarn at create/list time when the scheduler is not ready; stay silent on an unknown probe result.FNget_active_profile_nameu+ ⚠ Scheduler is not ready for profile 'z6': no gateway serves it and no fresh ticker heartbeat.z If no gateway is running: hermes gateway install sudo hermes gateway install --system # Linux servers Check status: hermes cron status)rEhermes_cli.profilesrJprintr r YELLOWDIMrIs r_warn_if_gateway_not_runningrOLsc "%/; %=>U>W=XYFFGM}} VW %9:@** FGrc0V^8dQhR\R\/#)rsecondsr)floatr)rs"rrrZs D De D Drc8\^\V44pT^<8dT R2#\ T^<4wr\ T^<4wr1\ T^4wrCTR3TR3T'gTM^R3.pRP RT44;'gR# \\3dR#i;i) z=Render a lateness duration compactly: '31m', '2h 30m', '45s'.?sdhm c3F"TFwrV'gKV V 2xK R#5iNr#).0nunits& r #_format_lateness..fs;WQLqc$Ls !!0m)maxint TypeError ValueErrordivmodjoin)rQminutes_hoursdayspartss& r_format_latenessrmZsaW&|!}$JGGR(NE#KDC[5#,!S(I JE 88;; ; C CtC z "sBBBc:V^8dQhR\\,/#r+)r r)rs"rrrisPP(3-Prc,RRRR/PV4#)catch_upzcatch-up after missed firelateget)kinds&r_dispatch_kind_labelruis 4ff E I I$ OOrcFV^8dQhR\R\\,/#)r next_run_atr)rr rR)rs"rrrms" H H3 H8E? Hrc^RIHp^RIHpV!V4pVfR#V!4P \ 4VP \ 4, P 4#)a8Seconds the stored ``next_run_at`` is already in the past (negative while still upcoming); None when it is not a parseable ISO timestamp. Parses through the scheduler's own ``_parse_aware`` so the CLI and the ticker agree on the instant (mixed UTC offsets, DST folds, legacy naive stamps read as system-local).  _parse_aware)nowN)rCrz hermes_timer{ astimezoner total_seconds)rwrzr{dts& r_next_run_overdue_secondsrmsF' k "B z E  S !BMM#$6 6 E E GGrctV^8dQhR\\\3,R\\\3,/#rjobr)dictrrtuple)rs"rrr}s. Q QtCH~ Q%S/ QrcVPRR4p\V4pVe:V\8:g/VPRR4'dVPR4R 9dRV3#R\V R\ V4 R 2\ P 43#) z``("Next run" | "Overdue", value)`` for one job. A stamp parked past `cron doctor`'s grace on a job that is supposed to fire is the only user-visible trace of a dead scheduler; never present it as an upcoming run (#114309). rwrTenabledTstatezNext runOverdue (u: ago — the job has not fired; is the scheduler running?)>paused completed)rsr_OVERDUE_GRACE_SECONDSr rmr rM)rstamp overdue_ss& r _next_run_rowr}s GGM3 'E)%0IY*@@779d++swww/?CZ/ZE"" uwc*:9*E)FG@@AGP QQrcFV^8dQhR\R\\,/#)rdispatchr)rr r)rs"rrrs:::#:rca\S\4'gR#V3RlR4wrpV'dV'd V'gR#\SPR^44pVR8Xd\ RV R2\ P 4#\V4;'gRp\ RV R 2\ P4R V R V R 2,\ R V R2\ P4,#)zOne-line scheduled-vs-actual dispatch summary; None when the stamp is malformed. On-time dispatches render dim; late/catch-up dispatches render loudly so a run fired long after gateway downtime doesn't look like an ordinary success. See #99879. Nc3F<"TFpSPV4xK R#5ir[rr)r\krs& rr_$_dispatch_display..sb8a1x||A8as!lateness_secondson_timezon time (scheduled )rq⚠ : z scheduled , ran rY( late)) scheduled_at dispatched_atrt) isinstancerrmrsr r rNrurM)r scheduledactualrtlatenesslabelsf r_dispatch_displayrs h % %b8abIt F$ -? CDH y*9+Q7DD  & 0 0&E Dr"FMM 2z)FSYRZZ[5\ \az(&--8 9:rc(V^8dQhR\RR/#)rtitlerNr,)rs"rrrs    rc\4RpRV R2RVP^04,R,RV R23F&p\\V\P44K( \4R#) zIBoxed cyan section header shared by ``cron list`` and ``cron incidents``.u┌u┐u│u└u┘Nu─────────────────────────────────────────────────────────────────────────u│ )rLljustr r CYAN)rrulelines& r _print_bannerrs] G DtfC"2U[[_"Du"LPSTXSYY\o^ eD&++&'_ Grc$V^8dQhR\/#r+r,)rs"rrrs``s`rc\V\4'd RP\\V44#\ V4#), )rrrgmapr) unverifieds&r_unverified_targetsrs0.8T.J.J499Sj) *_PST^P__rrz[paused]rz [completed]Fc$V^8dQhR\/#)rshow_allr6)rs"rrrs###rc v^RIHp\RRR7wr#VfR#V'g9VUu.uF,pVPRR4'gV!V4R8XgK*VNK. pp^RIHpV'g/\ \RV!4 R 2\P44R#\R V!4 R 24VFp\PV!V44;'g=VPRR4'dR \P3MR \P3p\ R\VPRR4\P4 R\V! 24\V4Fwrx\ RVR,R V 24K \!V4Fp \ RV 24K \ 4K \#4R#uupi)zList all scheduled jobs.)effective_job_stateT)include_disabledbriefNrrrIzNo scheduled jobs in profile 'zI'. Create one with 'hermes cron create ...' or the /cron command in chat.zScheduled Jobs (profile: rz[active]z [disabled] idrTrY :z<11)rCr_probe_then_list_jobsrsrKrJrLr r rNr _STATE_BADGESGREENREDrM _job_rows _job_warningsrO) rr _store_reportjobsrrJbadgervaluers & r cron_listrs-/TRM |  Cwwy$''+>s+Cx+O C4  <  e45L5N4OP]]^d^h^hj k-.E.G-HJK!!"5c":;dd*-'')T*B*BZ &W]WaWaHb  5s+V]];crc*VR,pVR8Xd\R\P4#VR8Xd\R\P4#VR8XdB\R\ VP R44 R\ V4 2\P4#\V RVP R R 4 2\P4p\VP R 4;'g^4pV^8d&V\R V R 2\P4, pV#) last_statusokdelivery_queuedz'finished; delivery is still in progressdelivery_failedz'ran, but the result was not delivered (last_delivery_error). r last_errorrTfailure_streakrz failures in a row)) r r rrM _short_reasonrs_delivery_fix_hintrrc)rrdisplaystreaks& r_last_run_displayrsm$KdT6<<((''> NN''>}SWWUjMk?l>mmp*3/0239==B B{m2cgglC&@%ABFJJOG )*//a 0F {53vh&9:FJJGG NrcV^8dQhR\\\3,R\\\\3,,/#r)rrrrr)rs"rrrs/"?"?4S>"?d5c?&;"?rcVPR4;'g/pVPR4pVPR4;'gR.pVPR4;'g$VPR4'd VR,.M.pVPR4;'gVPR4pVPR 4;'g/pVPR 4;'g/pR V'dR PV4MR 3RVPR43RV'dV R2MR 3RV'dVPR4MR 3RVPR4'd"\R\P4R,MR 3RVPR43RVPR43RVPR4'd!VPRR4 R \ V4 2MR 3R!\ VPR"443R#V'd(VPR$R4 R VPR%R4 2MR 3. pR&VPR'R(43R)VPR*VPR+/4PR,R443R-V'dVPR.^4 R/V 2MR03\V4R1\V\4'dTMR PV43.VU U u.uFwrV 'gKW3NK up p ,#uup p i)2z<``(label, value)`` detail rows for one job in ``cron list``.repeattimesdeliverlocalrskillmonitor_script monitor_url monitor_statelatest_executionSkillsrrScriptscriptMonitorz# (agent runs only on output change)Changedlast_changed_atModeno_agentzno-agentz# (script stdout delivered directly)WorkdirworkdirPython interpreterzLast runr last_run_atrTrDispatch last_dispatch ExecutionstatusrNamer2 (unnamed)Scheduleschedule_displayschedulerRepeatr/u∞Deliver) rsrgr r rNrrrrr) r repeat_info repeat_timesrrmonitor_source mon_stateroptionalrrs & rrrs''(#))rK??7+Lggi --WIG WWX  N NSWWW5E5ECL>2FWW-.HH#''-2HN(..BIww1288b 499V$B7 3778$% n~&&IJ  IMM"34BO  GGJ  z6::.1VV%' ) CGGI&' 377=)*  GGM " " s34B7H7M6NO(* , &sww'?@A  *--h<=R@P@T@TUY[^@_?`a"$ &'H -. SWW/R1H1L1LWVY1Z[\ ,kook156a ~F\abc z'377GTYYw=OP   -5>HLE%H>  ?? ?s K+K+c<V^8dQhR\R\R\/#)rrlimitr)rrcr)rs"rrr s&^^^C^#^rc"\T;'gR4P4P4pV'd'RPV^,P 44MRp\ V4V8dVRV^, R,#T;'gR#)zVFirst line of an adapter/error blob, whitespace-collapsed and capped, or 'no details'.rrYNu…z no details)rr splitlinesrgsplitlen)rrfirstreasons&& rrr so  O ! ! # . . 0E+0SXXeAhnn& 'bF,/K%,?F;UQY % ']fF\F\P\]rcRV^8dQhR\\\3,R\/#rr)rs"rrrs&OODcNOsOrc.RVPRR4 R2#)zOCheck the target with `hermes cron status` or change it with `hermes cron edit rz --deliver `.rrrs&rrrs'!!$v!6 77L NOrcV^8dQhR\\\3,R\\\3,R\/#)rrfire_errrr)rs"rrrs6IIDcNId38nIIrc zRVPRR4 R\VR,4 RVPRR4 R2#) zmissed scheduled fire at atrTrdetailz]. The messaging gateway was unreachable. Run `hermes gateway restart`, then `hermes cron run rrz` to run it now.)rsr)rrs&&r_missed_fire_issuersN' T3(?'@=QYZbQcCdBef #f 566F HIrchV^8dQhR\\\3,R\\,/#rrrrr)rs"rrrs&tCH~$s)rc h.pVPR4;p'dVPRV 24VPR4'dKVP\R\P4 R\ VR,4 R\ V4 24VPR4;p'd9VP\R\P4 R \V4 R 24VPR 4p\V\4'dMVPR 4'd6VP\R \W4 2\P44VPR4p\V\4'dVPR 4'd\VPR4;'gR4R9pTP\RVPRR4 R\ VR ,4 2V'dRMR,V'd\PM\P44V#)z;Delivery / fire warning lines for one job in ``cron list``.last_delivery_queuedzNDelivery still in progress (the result was handed off but not confirmed yet): ru!⚠ The result was not delivered:rYz. last_delivery_unverifiedu⚠ Delivery UNVERIFIED:z adapter acked z without message_id/raw_responselast_fire_errorrr last_failurerru⚠ Last failure at rrTru — recovered since>rr) rsrr r rMrrrrrrrr)rlinesqueuedrrr recovereds& rrrs E/00v0 eflemno ww$%% A6==QRRS%c*?&@AB"EWX[E\D]_ `WW788z8 8&--HI+J788XZ [ww()H(D!!hll8&<&< UT"4S"C!DEvzzRS 77>*L,%%,*:*:8*D*D .44"59RR  U"<#3#3D##>"?r\(345 7T]:Pce g&FMMFJJ8 9 Lrc^^RIHpHp^RIHpV!R4V!RRR7^# Td=p\ \ RT 2\P44\ R4Rp?^#Rp?i\d=p\ \ R T 2\P44\ R 4Rp?^#Rp?ii;i) zRun due jobs once and exit (external-scheduler mode: a system crontab calls this). No gateway runs here by definition, and none is started: agent jobs are skipped with a logged reason until the gateway is up (headless surfaces refuse rather than spawn).)CronTickYieldedtick) begin_processcronT)verboseheadlessu✗ zn A fresher gateway process owns the runtime lock and will fire due jobs; this stale process yielded its tick.Nu✗ Cron tick failed: z= Check `hermes cron status` and the gateway log for details.) cron.schedulerr!r"/hermes_cli.observability.shared_metrics_processr#rLr r rMOSErrorr)r!r"r#excs r cron_tickr+9s 5M& TD)   ed3%L&--01 0 1  e,SE2FJJ?@ MN s& $B,1A"" B,/B,01B''B,cFV^8dQhR\\,R\/#)rjob_idr)r rrc)rs"rrrTs , ,hsm ,3 ,rc^RIHpV!WR7pV'g\R4R#VFp\VPRR4 RVPRR4R R VPR R4 R VPR R4 RVPRR4 2 4VPR4'gK\RVR, 24K R#)z,Show indexed durable cron execution history.)list_executions)r-rz$No cron execution attempts recorded.NrrTrrz<9z job=r-z source=source claimed_aterrorr)cron.executionsr/rLrs)r-rr/recordsrecords&& r cron_runsr6Ts/V9G  45 D#&'r&**Xs*CB)GHZZ#./yHc9R8SSUL#./1 2 ::g   D)* + rdetectedalertedresolvedclosedc$V^8dQhR\/#r+rc)rs"rrrgs+ + C+ rc ^RIHpHp\VRR4pVR8Xd\VRR4pV'g&\ \ R\ P44^#V!V4'd*\ \ RV R 2\ P44^#\ \ R V R 2\ P44^#\VR R4pV!VR 7pV'gU\ \ R\ P44V'd(\ \ RV R2\ P44^#\R4VEFup\ VR ,\PVR ,\ P44p\P!RRVPR4;'gR4P!4p \#V 4^8dV R,P%4R,p RVR,3RVPRR43RVPRR43R VPR!R43R"V 3R#VPR$43.p \ R%\ VR&,\ P4 R%V 24V F.wrV R#8wg V 'gK\ R'V R(,R) V 24K0 \ 4EKx \ \ R%\#V4 R*2\ P44^#)+zList (``[--state ]``) or ``ack `` durable cron failure incidents. Acking closes an incident so its failure ping stays silent until the error signature changes. ) ack_incidentlist_incidentsincident_actionrack incident_idNuA✗ Incident ID required: hermes cron incidents ack u ✓ Incident z acknowledged (closed).z Incident z not found or already closed.r)rz#No cron failure incidents recorded.z (filtered by state 'z')zCron Failure Incidentsz\s+rYr2r:NNz...Jobr-Type failure_typeunknownz First seen first_seen_atrTz Last seen last_seen_atErrorOutput output_filerrrrz<12z= incident(s) | ack one with: hermes cron incidents ack )cron.incidentsr>r?getattrrLr r rrrMrNr_INCIDENT_STATE_COLORSrsresubrr rstrip) argsr>r?actionrBr incidentsinc state_display error_textrowsrrs & rcron_incidentsrZgsg < T,f 5F dM48  %[]c]g]gh i  $ $ %- }4KLfll[ \ %)K=0MNPVP]P]^ _ D'4 (EU+I  e96::FG  %0r:FJJG H*+c'l,B,F,Fs7|U[U_U_,`a VVFC)9)?)?R@FFH z?S #D)002U:JH &1S(Tsww<=cggnc:;gz=R377=124 5TFMM232m_EF LE EEUS[-eW56!  %"S^$$ab   rz Hint: jobs.json may be owned by another user (e.g. rewritten by a root `docker exec hermes hermes cron ...`). Fix ownership to match the gateway user, and prefer `docker exec -u :`.z Hint: the ticker hit file-descriptor exhaustion (EMFILE). The scheduler now retries with backoff and attempts fd reclamation, but if the leak persists, restart the gateway to recover scheduling.cFV^8dQhR\\,R\/#)rager)r rRr6)rs"rrrs"GGhuoG$GrcJ^RIHpVRJ;'dW^,^,8*#)r!)TICKER_INTERVAL_SECONDSN)rCr^)r\r^s& rrDrDs$1 d? F Fs&AB&FFFrc:V^8dQhR\\,/#r+)r r)rs"rrrs88(4.8rcH^RIHp^RIHpV!V!4P4#)zProbe the active cron store (never a marker inside it): ``None`` when it accepts writes. Runs before any job read: loading jobs re-secures the dir, which can mask a mode problem.)_current_cron_store) probe_report)rCracron.store_healthrbcron_dir)rarbs r_store_unwritable_reportres.. +-66 77rrc0V^8dQhR\R\/#)rrrr)rs"rrrs  t D rc^RIHp\4;pe \W1R7W2!VR73# \d TfhTR3u#i;i)aI``(store report or None, jobs or None)``. Probes and prints the unwritable-store warning BEFORE reading jobs: loading re-secures the dir mode (masking a mode problem), and on an unwritable store the read itself can fail (it creates cron/output). Jobs is ``None`` only when that read failed on a store already reported unwritable. list_jobsN)rr)rCrire_print_store_unwritabler))rrrireports$$ rrrsT $*,,94y2BCCC  > t|s ,AAc4V^8dQhR\R\RR/#)rrlrrN)rr6)rs"rrrs&kkDkDkTkrc V'dE\\RVR, RVR, RVR, R2\P44R#\\R\P44\\R VR, R VR, R VR , R 2\P44\\RVR, R2\P44R#)u⚠ Cron store storez is NOT writable (r2u() — scheduled jobs are being skipped; fixz. See `hermes cron status`.NuC⚠ Cron store is NOT writable — scheduled jobs are being skippedrrz (last successful write sincerz Fix: z0. Each due job then fires once on the next tick.)rLr r rrM)rlrs&$rrkrks  eofWo%66HPWHYZ**0-8SUV\V`V`b c %UW]WaWa bc %"VG_%Rw'88PQWX_Q`Paabcekeoeo pq %'&-(XY[a[h[h ijrc4V^8dQhR\R\RR/#)rpidsrestart_commandrN)rr)rs"rrrs"<=<=t<=c<=Y]<=rc a ^RIHpHpHp^RIHp^RIHpV!4pV!4pV!4p V'd#RRP\\V44 2MRo RV 3Rllp VfV !R 4\R V 24R#\V4'g&V !R \V4 R 24\R V 24R#V!V 4;p eQV !R4\\RV ^, RV ^, R2\P44\RV 24R#Ve\V4'd VfV 'dT !RVe#R\\V44,R,MR R24V 'd\\RV 2\P44RV 9d)\\\ \P"44M5V!V 4'd(\\\$\P"44\R4R#\\R\P&44S 'd \S 4Ve\R\V4 R24R#R#)zReport builtin-ticker liveness for a gateway process known to be alive. The ticker THREAD can die silently or stay alive while every tick fails, so check both the liveness heartbeat and the last-successful-tick marker before saying "will fire". )r=get_ticker_last_errorget_ticker_success_age)_is_fd_exhaustion_text)stale_code_yield_labelsz PID: rNc(V^8dQhR\RR/#)rheadlinerNr,)rs"rr*_print_ticker_health..__annotate__srcx<\\V\P44S'd\S4R#R#r[)rLr r rM)r{pid_lines&r_warn#_print_ticker_health.._warns% eHfmm,-  (O ruH⚠ Gateway is running but the cron ticker has not reported a heartbeat.z Cron jobs will NOT fire until the ticker writes its first heartbeat. If the gateway just started, wait ~60s and re-run `hermes cron status`. If heartbeat never appears, restart: uN⚠ Gateway is running but the cron ticker looks STALLED — no heartbeat for zs (expected every ~60s).z( Cron jobs may NOT be firing. Restart: uZ⚠ Gateway is running STALE code — its cron ticker yields every tick and fires NOTHING.z Booted on z, checkout is now at z2 (the code was updated under the running gateway).z Restart it onto the new code: uC⚠ Gateway and cron ticker are running, but no tick has succeeded zin rUyetu — ticks may be failing.z Last tick error: zPermission deniedz. Check the gateway log for 'Cron tick error'.u<✓ Gateway is running — cron jobs will fire automaticallyz Ticker heartbeat: zs ago)rCr=rvrwr'rxryrgrrrLrDrcr r r_PERMISSION_HINTrM_FD_EXHAUSTION_HINTr) rsrtr=rvrw_cron_is_fd_exhaustion_textryhb_ageok_agerrskewr~s && @r_print_ticker_healthrs"QQT6 % 'F # %F&(J8<3sD>234$H ~ XY 88G7HJ K"& ) ) ""%f+.FH I 88IJK)*5 5$ B   el47)+@a JHHIOU V 00ABC  %9&%A%Av~Zd =C=O53s6{#33c9UZ[\** +  %-j\:FJJG H"j0e,fmm<=,Z88e/?@ >? eRTZT`T`ab  (O   (V U; < rc ^RIHpHp^RIHp\ 4\ RR7wr4Vf%\ \R\P44M=Ve:\RV44p\ \RV R 2\P44Ve \ 4\4pVR 8wdM\ \R V R 2\P44\ \R \P44EM^RIHpV!4pRp \ P"!\$4;_uu_4V!V4p RRR4V e.MV!4p Rp Rp Rp V fV 'g\ P"!\$4;_uu_4^RIHpHpV!4p V 'dV!4MRpV'dV.MT p RRR4V 'gV!4p V 'g[V 'gS\ P"!\$4;_uu_4^RIHpHp\3V!44;'dV!4p RRR4V e\ RV P54 24Rp\ P"!\$4;_uu_4^RIHpV!V4pRRR4Ve VP8MRp\;V P<.RV R2R7EM9V 'gV 'gV 'g V 'dSV 'd\ R4\;.RR7MV 'd\ RV R24\;.RR7M\;V 4M\ \RV R2\P44\ P"!\$4;_uu_4^RIHpHpV!4pVeFVV^,^,8d1\ \R\AV4 R 2\PB44RRR4\ R!4VR#9d \ R"4\ 4\EV4\ 4R# +'giEL;i +'giEL{;i +'giEL;i +'giEL;i +'giL;i)$zShow cron execution status.r9rIFrjNz3 Jobs could not be read from the unwritable store.c3|"TF2p\VPR44;'g^^8gK.^xK4 R#5i)rwN)rrs)r\js& rr_cron_status.. s3kA1J155Q^K_1`1e1edeij0jaas %<< <rz due run(s) not firedr0u✓ Cron provider: uD — jobs fire via the managed scheduler, not the in-process ticker.zq (No ticker heartbeat is expected for an external provider; due jobs are delivered by an authenticated webhook.))host_gateway_serving)get_running_pidr8r<z Scheduler host: )host_multiplexer_servingdefaultzhermes --profile z gateway restart)rtz> Scheduler host: the host gateway (multiplexing this profile)z(hermes --profile default gateway restartzY Scheduler host: an in-process ticker (hermes serve / Desktop backend) ticking profile ''z5restart the Hermes Desktop app (or its serve backend)u%✗ No scheduler is serving profile 'u!' — its cron jobs will NOT fire)r^r=z Scheduler last ticked u6 ago — jobs that came due since then have not fired.a3 Start the ONE host gateway (it multiplexes every profile, this one included): hermes --profile default gateway install # user service sudo hermes --profile default gateway install --system # Linux servers: boot-time service hermes --profile default gateway run # Or run in foregroundak It serves this profile automatically. If a per-profile service or gateway from an older release is still installed, fold it in (preflight + dry run): hermes --profile default gateway migrate --multiplex --dry-run hermes --profile default gateway migrate --multiplex Check: hermes cron status from this profile should show its ticker heartbeat. )rcustom)#rBr:r;rKrJrLrr r rsumr4rrNgateway.host_topologyrr?r@r3rArr8rCr=r>rDdescriber profile_labelrpidr^rmrM_print_active_jobs_summary)r:r;rJ store_report active_jobsoverdueproviderractivehostrsgateway_alive_via_lockserved_by_multiplexerin_process_tickerrr8lock_pidr=r>ownerr host_profiler^rs r cron_statusrsa; G 5u ML eI6::VW  !kkk eb !67DE )+H9 e)(41128,,@ A eKLRJJX Y ?(*   + +'/D,%r+<+>!& %! <$$Y//[)G)I&0F?,D%-z40*(S(U%)2G ((33a)=>V>X)Y*M*M-J-L&4   &t}}&78 9E$$Y//G080382C5..L $((?PQ]P^^n=o p +/DHY$VW$R9cd"qrxqyyz{|$R9pq$T* %?xGhikqkukuv w$$Y//W13%&3JQ3NQS3S*S% :#3F#;"<==!=>DmmMN 0 Z [22jk  G{+ GS, + +0//4330//(0/sC( P:(P#P1(P.%Q'AQ P  P+ . P?  Q  Q& cV^8dQhRR/#rGr#)rs"rrrls&?&?&?rcV'g\R4R#^RIHpVUu.uF>pV!VPR44;pfK VP \ 4VR,3NK@ pp\R\ V4 R24V'dr\VRR7^,p\V4pVe@V\8d5\\R V R \V4 R 2\P44M\R V 24VUu.uFKp\VPR 4\4'gK*VR ,PR4R9gKIVNKM ppV'Ed\4\\R\ V4 R2\P44VFpVR ,p\VPR^44p \RVPRR4 RVPRR4 R\!VPR44 RVPRR4 RVPRR4 R2 \RV R2\P4,4K R#R#uupiuupi) zJPrint the ' active job(s)' + next-run line shared by every status path.z No active jobsNryrwrz active job(s)cV^,#r!r#)runs&r,_print_active_jobs_summary..{s#a&r)keyu ⚠ Next run u is OVERDUE — passed z: ago but the job has not fired (is the scheduler running?) Next run: rrtu ⚠ z job(s) last fired late:rrrrTr2rrz , scheduled rrrrYrr)rqrp)rLrCrzrsr}rr minrrr rmr rMrrru) rrzrparsed next_runsearliest overdue_byrqrVlate_bys & rrrls4   !& FJPT+AEE-,@AAV<&##C(!M*:;TIP Bs4yk ()y&89!<.x8  !j3I&I %/(3J+J789667=}}F G L + , Ht!z!%%*@$GA/"&&v.2FF AtD H t  efSYK'?@&--PQA/"A&quu-?'CDG DtS)*"QUU6;-G,HK_`a`e`efl`mKnJop uu^S9:&X[A\@]]^`AgYf-v}}=> ? 'P" HsI"I;'I'IIc$V^8dQhR\/#r+r)rs"rrrs''t'rc4^RIHpVPR, #)uYScripts dir for cron jobs — via ``CRON_DIR`` so monkeypatched cron storage is honoured.CRON_DIRscripts)rCrparentrs r_scripts_dir_for_cronrs" ??Y &&rcFV^8dQhR\R\\,/#)rrrrr )rs"rrrs   # rc\4P4p\V4P4pVP 4'dVP4MW, P4pVP V4TP4'gRT 2#TP4'gRT 2#R# \ d RT RT: 2u#i;i)z=Human-readable script issue, or ``None`` when the path is OK.zscript resolves outside rzscript not found: zscript path is not a file: N) rresolver expanduser is_absolute relative_toreexistsis_file)r scripts_dirrawpaths& r_script_health_issuers')113K v, ! ! #COO--3;;=K4E3N3N3PDD % ;;==#D6** <<>>,TF33  D)+b CCDs/B88CCcFV^8dQhR\R\\,/#)rnext_runrr)rs"rrrs"``c`hsm`rc\V4pVfRV: 2#V\8:dR#VR8dVR, R R2M V^<, R R2pRV R 2#) z8Issue string when ``next_run_at`` is parked in the past.Nz&next_run_at is not a valid timestamp: iz.1frWz.0frXznext_run_at is u: overdue — job is not firing (is the scheduler running?))rr)rramounts& r_next_run_overdue_issuersi)(3I7|DD**+4+< D %Q 'YQS^TWDXXYBZF VH$^ __rchV^8dQhR\\\3,R\\,/#rr)rs"rrrs&""T#s(^"S "rc b.p\VPR4;'gR4P4P4pV'dLVR9dETP R\VPR4;'gR4P4 24\VPR4;'gR4P4;p'd*VP R\ V4 R\ V4 24VPR4;p'd"VP R \V4,4\VPR 4;p\4'dg\VPR 44;p'dEVP R V R VPRR4 R\VPR^44 R24\VPR4;p\4'd2VPR4'dVP \W44VPRR4'dwVPR4R 9da\VPR4;'gR4P4pV'd \V4MRp V 'dVP V 4\VPR4;'gR4P4p VPR4'dV 'gVP R4V 'd%\V 4;p 'dVP V 4\VPR4;'gR4P4p V 'dB\V 4P!4P#4'gVP RV 24V#)!rrzlast run failed: r unknown errorrz4last run finished but the result was not delivered (rrz;last delivery unverified (adapter acked without evidence): rrtzlast fire was z (scheduled rrTrrz5 late). This warning clears at the next on-time fire.rrrTrrwzactive job has no next_run_atrrzno-agent job has no scriptrzworkdir not found: >rrr>rr)rrsrlowerrrrrrrrurmrrrrrr) rissuesr delivery_errrrrrrissuer script_issuers & r_cron_doctor_issues_for_jobrsFcggm,22399;AACK{*VV )#cggl.C.V.V*W*]*]*_)`ab377#89??R@FFHH|H L][gMhLiil+C013 4WW788z8 S+J78 9cggo66(==(f)=> >5 > MMN5'hll>[^>_=``b-hll;Mq.QRSTJJ Kcgg&788($??HLLQYDZDZ (78 wwy$CGGG$4.s<8iac&kk8szCron doctor found z issue(s) across z job(s):rrrTrYr2rz - z?Review the findings above, then run `hermes cron doctor` again.) rCrirrLr r rr rNrrMrs)rirrrfindingsnote issue_countrs r cron_doctorrsA# e ,D)-^#CI;o6Db eD&**%&<8<Z=[\]E F5'" #  G %QSYS]S] ^_ _s F F cFV^8dQhR\\\3,/#r+r)rs"rrrsUUT#s(^Urc X\UUu/uFwrV\WR4bK upp#uuppi)zMCollect the create/update kwargs shared by ``cron create`` and ``cron edit``.N)_JOB_ARG_FIELDSrN)rSapi_keyattrs& r_job_api_kwargsrs(DS TO=7GWT. .O TT Ts&cJV^8dQhR\\\3,RR/#)rjob_datarNr)rs"rrrs"22c3h2D2rc\F>wrVPV4'gK\VPW,44K@ R#)zPPrint the optional Script/Monitor/Mode/Continuity/Workdir lines of a job record.N)_JOB_DETAIL_LINESrsrLr)rrtemplates& r_print_job_detailsrs0*  <<   (//(-0 1+rc\RRRRVPRVPR\VRR4R\ \VRR4\VRR44R\VRR4;'gR/\VR R4'g\VR R4eR VP R \VR R4/M/B\ V4BpVPR 4'g9\\R VPR R4 2\P44^#\\RVR, 2\P44\RVR, RVR, 24VPR4'd%\RRPVR,4 24\VPR/44VPR/4PRR4'g \R4M\RVR, 24\4^#)rTcreaterpromptrNrrFr paused_reasonsuccesszFailed to create job: r2rz Created job: r- Name: r2 Schedule: Skills: rrrTu? Created PAUSED — resume to schedule, or explicitly run now.rrwr#)r)rrrNrrrrsrLr r rrrgrrO)rSresults& r cron_creaters!!"&--!8< !dGT*!!w!=wtXW[?\]!z5199T ! dHe,,ot0T0` dkk?GD/SW4X Yfh ! $ !F ::i  e,VZZ-Q,RSU[U_U_`a %-x 012FLL AB HVF^$N6*3E2F GH zz( 499VH%56789vzz%,- ::eR $ $Y 5 5 OP VM2345 " rc`^RIHpHpV!VP4pT'g3\ \ RTP 2\P44^#\TPR 4;'g$TPR 4'd TR ,.M.4p\\TR R4\TR R44p\R\TR R44;'g.p\\R\TR R44;'g.4p Rp \TR R4'd.p MPTeTp MIT'g T 'd:TU u.uF qT 9gK T NK p p YU u.uF qT 9gK T NK up , p \R RRRTPR\TRR4R\TRR4R T R\TRR4/\!T4Bp T PR4'g9\ \ RT PRR4 2\P44^#T R,p \ \ RT R, 2\P"44\ RT R, RT R, 24\ T PR 4'dRRP%T R ,4 2MR4\'T 4^# Tdtp\ \ \ T4\P44TPF,p\ RTR, RTPR4: R24K. Rp?^#Rp?ii;iuup iuup i)!r!)AmbiguousJobReferenceresolve_job_refrrz (name: r2rNJob not found: rr add_skills remove_skills clear_skillsFrTupdater-rrrrzFailed to update job: r2rrz Updated job: rrrrz Skills: noner#)rCrrr-rLr rr rmatchesrsrrrNsetr)rrrgr)rSrrrr*rXexisting_skillsreplacement_skillsrr final_skillsrrupdateds& r cron_editr1s@dkk*  eodkk]3VZZ@A3778,\\3777CSCS#g,Y[]O*74$+GQUW_aeIfg"4|T)JKQQrJ)$ot0TU[[Y[\MLt^U++  ') }+:Y?%=>X? YJTJ5|:SJTT ZhZt{{Z 'j$ ?Z%dHd;ZDPZ!(j$ ?ZDSSWCXZF ::i  e,VZZ-Q,RSU[U_U_`aUmG %- 123V\\ BC HWV_%^GJ4G3H IJ X9N9NJtyy!234 5 w E ! eCHfjj)*A Bqwiyv(9; < &ZTs0J);L&L&L+!L+)L#0A(LL#cHV^8dQhR\R\R\R\/#)rrTr- success_verbr)rrc)rs"rrrZs(   S   rc hRpVR8XdC\P!\4;_uu_4^RIHpVP R4pRRR4\ WR7pVeXPV4TPR4'g<\\RT RTPRR 4 2\P44^#TPR 4;'gTPR 4;'g/p\\T RTPR T4 R T R2\P44TR9dDTPR /4PR4'd\RTR ,R, 24TR8Xd\R\TPR /44 24TPR4'dETPR4RJd0TPR4'gTPR4R8wd^#^# +'giEL;i TeXPT4ii;i)Nr)_SESSION_ASYNC_DELIVERYF)rTr-rz Failed to z job: r2rr removed_jobr2 (rrwrrexecutedexecution_success delegation_idexecution_mode background>rresume)r?r@r3gateway.session_contextrrr)resetrsrLr r rr _run_outcome)rTr-r_stateless_tokenrrrs&&& r _job_actionrZs    + + H6::5A  ,<&8  ' # ) )*: ; ::i  ejvzz'?/S.TUW]WaWabc **U  > >vzz-8 > >BC %<.swwvv'>&?r&KV\\ Z[ ""vzz%'<'@'@'O'O VE]=9:;<  < 5" 56789 GGJ  CGG,?$@E$I00SWW=M5NR^5^ 1, + +  ' # ) )*: ; (sH H H H1cRV^8dQhR\\\3,R\/#rr)rs"rrr|s& U Ud38n U Urc"VPR4'dRVR, R2#VPR4R8XdR#VPR4'd#RVPR 4'dR R 2#R R 2#VPR 4;'gR#)zOne-line verdict for a manual run. A background-dispatched run (execution_mode="background" / delegation_id) keeps running after this CLI exits, so report the dispatch rather than a success/failure verdict. r z"Running in background (delegation z).r rzRunning in background.r z Ran now: r  succeededfailed.execution_skippedz'It will run on the next scheduler tick.rrrs&rrr|s  ww3C4H3ILL ww L0' wwz#''2E*F*F;UUVWWHUUVWW 77& ' T T+TTrc$V^8dQhR\/#r+r<)rs"rrrs   rc ,\VRR4p\VRR4pV'd.V'd&\\R\P44^#V'g V'g\ RVP R4#^RIHpH pH pT!VP V'dV!4P4MT4pT'g3\\R TP 2\P44^#\\R TPR TP 4 R TP R2\P4RTPR4 2,4^# T\3d2p\\R T 2\P44Rp?^#Rp?ii;i)z>Resume a paused job or explicitly re-arm a completed one-shot.run_atNrun_nowFz%Use exactly one of --at or --run-now.rResumed)r _hermes_now rearm_oneshotzFailed to re-arm job: rzRe-armed job: r2r rz Next run: rw)rNrLr r rrr-rCrr r! isoformatrersr)rSrrrr r!rr*s& r cron_resumer#s/ T8T *FdIu-G ' e;VZZHI '8T[[)<<KKDKKg)@)@)BSYZ  eodkk]3VZZ@A %.!= >b QOQWQ]Q] ^SWW]34 5 67  ": . e,SE2FJJ?@sEE F"&FFc$V^8dQhR\/#r+r<)rs"rrrs++#+rcF^RIHp\\VRR4;'gR4p\VRR4;'gRp\VRR4p\VRR4pV'g&\ \ R \ P44^#VR9dVPV4pV'g(\ \ R V R 2\ P44VFlp\ R\ VR,\ P4 RVR, R\ R\VR,4,\ P4 24Kn ^#VR 8XdRMV R2pVe VR 8Xd-Vf)\ \ RV 2\ P44^#VR 8Xd?VPW$V4\ \ RV RV R2\ P44^#VR 8Xd$VPW$4p V e\ V 4^#MCVPW$4'd-\ \ RV RV R2\ P44^#\ \ RV RV R2\ P44^# \d2p \ \ RT 2\ P44Rp ? ^#Rp ? ii;i)zHandle ``hermes cron notepad [get|set|delete|list]`` (per-job durable KV). A running cron agent updates its own notepad via its terminal tool; the scheduler injects non-empty notepads into the job prompt on each run. )notepadr-rnotepad_actionNrrrzA job ID is required.rrszNotepad for job z is empty.rz = z z updated: updated_atzset z z$Usage: hermes cron notepad zSet notepad key 'z ' for job rzDeleted notepad key 'zNo notepad key 'zNotepad error: )rrsdelete)r$r&rrNrLr r r list_notesrNrMset_noterget_note delete_notere) rSr&r-rTrrnotesr usage_argsstoredr*s & r cron_notepadr1sH  x,22 3F T+T 2 < zlKVZZX Y U?   V% 0 %+C5 6(!DfllS T U?%%f2F!f "  - - %/uJvhaH&,,W X e&se:fXQ?OP  eocU+VZZ89s@I$&BI$AI$AI$ 'I$4AI$8*I$$ J /&JJ rcB\\VRR44;'g^#)allF)rrNas&rrrsi5% 89>>Q>rrc*\4;'g^#r)rr4s&rrrs ***rdoctorc\4#r[)rr4s&rrrs rr"c\4#r[)r+r4s&rrrsikrrunscZ\\VRR4\VR^44;'g^#)r-Nr)r6rNr4s&rrrs*i8T :GAwPR!,rr&c\V4#r[)r1r4s&rrrsarrc\V4#r[)rr4s&rrr Areditc\V4#r[)rr4s&rrrsilrpausec0\RVPR4#)rBPausedrr-r4s&rrrs{7AHHh?rrc\V4#r[)r#r4s&rrrr?rrc0\RVPR4#)r TriggeredrEr4s&rrrs[+>rremovec0\RVPR4#)rIRemovedrEr4s&rrrs Hahh Brhistoryaddrmr)c\VRR4p\PVfRMT4pVe V!V4#\RV R24\P !^4R#)zHandle cron subcommands. cron_commandNrzUnknown cron command: zV Usage: hermes cron [list|create|edit|pause|resume|run|remove|status|runs|doctor|tick])rN_CRON_SUBCOMMANDSrsrLsysexit)rSsubcmdhandlers& rrPrPs] T>4 0F##fnF&IGt} "6(+b bcHHQKr)NN)F)x)N)zhermes gateway restarti))r2r2)rr)failure_deliverrX)rr)rr)rr)modelrY)rmodel_provider)pinnedr[)rr)rr) continuityr\)reasoning_effortr])rr))rz Script: {})r0 Monitor: {} (agent runs only on output change))rr^)rz3 Mode: no-agent (script stdout delivered directly))r\z: Continuity: on (each run sees the previous run's output))rz Workdir: {})rz Python: {})R__doc__r?r%rPrRdatetimerrpathlibrtypingrrrr r __file__rr PROJECT_ROOTrinsertrhermes_cli.colorsr r rr)r4rErOrmrurrrrrrMBLUErrrrrrrrr+r6rrrNrOrZrrrDrerrkrrrrrrrrrrrrrrrrrr#r1rQrPr#rrrhsj% "66H~$$++335 3|$%+ . B G DP H Q:* `J 6 mU[U`U`Eab #D""?J^O I < 6 ,%fjj)V]]JX^XdXd"FJJ0+ \EN G 8 5  k5k<=~f R&?R' (!`"J ,3U %2 6& R D U .+` > * % ! Y, ( & " ? & > B 18),X6%8I(8SS$+H5r