+ jHj54Rt^RIt^RIt^RIt^RIHtHtHtHt]P!] 4t Rt Rt RtRtRtRtR tRR ltR tR tR tRtRtRtR#)zMFence an unsettled claim in a live owner without inventing a terminal result.N)RuntimeStoreError _admission_epochsettle_steer_transferscaaa\SSPPSR,4S4oVVV3RlpSPP V4#)target_session_idc<\VSP4\VSR,4pVR,SP8wgVR,SR,8wd \R4hVR,R 9d VR,#VR,R8wd \R4hVP RSR,34\ VSPSR,SRR7VP R SR ,34R#) admission_id owner_epoch generationstale_generationstatusunknownstartedzCUPDATE session_admissions SET status='unknown' WHERE admission_id=?)epochr transfersoutcomezBUPDATE sessions SET runtime_revision=runtime_revision+1 WHERE id=?r>rterminal)rrrrexecuter)conncurrent authorityrowrs& H/home/ubuntu/.hermes/hermes-agent/gateway/session_settlement_recovery.pywrite_mark_unknown..writestY__%T3~#67 = !Y__ 4 8MQTUaQb8b#$67 7 8  7 78$ $ 8  )#$67 7 Z]`ao]p\rst9??s?R')"4 "E;7S\K]; M=IMM I )*H ~~2**84::O!!/2G gt $ $();w)FG1 c/*=*=j*I*O*OR&Po))'288b9;'4  $+!O''(9::&*"#>**   3 4 4-1H) *   5 6 6/3H+ , o)))4c : :y?Y?Y"1)">r&&7>>3%h&?@/%%&67>>&J8H(I&J&5s<7H&J!" OrcVPPVPR4\ W4WV4p\WdR,V4pV'dVPR4'gWR&VPPVPV4R# \d!\P RTR,4Li;i)zThread-safe half of terminal publication; the caller holds the stream lock. No SQLite write happens here: loop-side readers wait on this lock, so it never spans storage I/O.Nz0Terminal control cleanup failed for admission %sr rrH) controlssnapshot session_id Exceptionlogger exceptionr_owner_stop_noticerr&publish) rr(refrrZr[r\r<notices &&&&&&& r_publish_completionrkhsb s~~t4"#CH x);S AF hll9--%cnnh7 bKSQ_M`abs&B(B=<B=zxGateway restarting: this turn was interrupted and will not be run again. Send your message again once Hermes reconnects.zxGateway shutting down: this turn was interrupted and will not be run again. Send your message again once Hermes is back.c^RIHpHpVR8Xd5T;'g/PR4;'g/PR4MRpV\V\ /PV4pVe!Ve^RIHpV!V4'd ^RIH pV#V#)aThe named notice for a turn the owner's restart/stop interrupted, or None. With the default ``agent.restart_drain_timeout: 0`` a restart interrupts in-flight turns at once; the admission settles ``interrupted`` and is never replayed, so the viewer is told the restart did it. An unattended ``row`` (cron, automation, webhook) has nobody to resend it, so it gets the neutral automation copy instead of resend advice (#134889 GW-20, GW-10's rule).)!_INTERRUPT_REASON_GATEWAY_RESTART"_INTERRUPT_REASON_GATEWAY_SHUTDOWNr9r>interrupt_messageN) is_unattended)AUTOMATION_INTERRUPTED_NOTICE) gateway.runrmrnrRESTART_INTERRUPT_NOTICESHUTDOWN_INTERRUPT_NOTICE"gateway.session_unattended_unknownrpagent.turn_failure_copyrq) r\rrrmrncauserjrprqs &&& rrgrg~sbOVZgOghnn" ! !( + 1 1r 6 67J KmqE/1I02KMMPSQVZ  coD    M0 0 Mrc*T;'g/PR4;'g/pVPR4VPR4rCVR8wg5\V\4'd\V\4'd V'gR#VRVR\V4/3#)z``(notice, display_metadata)`` for a failed turn the gateway itself refused with a renderable descriptor (``session_cron.execute``: the only producer of ``result.error_surface``), else None.r>r?r=failedN)rrTrUrW)r\rvaluer^r=s&& rgateway_refusal_boundaryr{sz^^  * 0 0bEYY/71CU(*Wd";";:eUXCYCYaf 7E?DMB BBrc^RIHpV!VPVR,4R# \d"\P RTR,4R#i;i)zYTerminal media cleanup: its own write txn after publication, never under the stream lock.)release_admission_mediar z.Terminal media cleanup failed for admission %sN)gateway.session_ingress_mediar}r rdrerf)rrr}s&& r_release_mediarsE`I c..AB `I3~K^_`s $(AAc ^RIHpHp V !W4WV4p VR,p VPP;_uu_4W1P V &RRR4\ WV4p V!VPVPW4VWjV R7wVR&VR&VPP;_uu_4VP PV R4\WW#VR,VR,V4RRR4VPP;_uu_4VP PV R4RRR4\Y4R# +'giL;i +'giLs;i +'giLC;i TPP;_uu_4TP PT R4RRR4i +'gii;i;i)aWorker-thread settlement: redaction, encoding and the SQLite write never stall the owner loop. The write runs OUTSIDE ``event_stream.lock``: attach, replay, receipts and pending publication take that lock on the loop, and a writer held by another process (up to ``_WRITE_PATIENCE_S``) would stall every session behind it. Ordering is kept by ``live.settling``: from before the write until the completion frame is published, lock holders read this claim as still started, so none sees the terminal row without its completion. ``settlement`` receives the committed ``(settled, response)`` before publication, so a publish failure is not a lost commit. The caller publishes idle ``session.info`` on the loop afterwards (bot receipt wakeups are tasks). ) finish_resultprepare_resultr N)rrr[rr>preparedrrZr[) gateway.session_resultsrrr&r'settlingrr rpoprkr) rr(rirr[rr\ settlementrrrr rs &&&&&&&& rcommit_and_publishrsSFcW?H~&L      &) l# 9C0I 28E LL SU\)9E5 9z*5   # # # MM  lD 1  :i;PR\]gRhjr s$   # # # MM  lD 1$9" $ #$ #T   # # # MM  lD 1$ # # #sTD8A E1&:E  E1E8 E E E1 E. 1G F6 - G 6 G G c"^RIHp.\ORNFEp \P!V!V\P !\ WW#WEWg4 44GRjxL u# R#L \\P3dUp T eRT9dh\PRTR,Y4\P!T 4GRjxL Rp ? KRp ? ii;i5i)aCommit the captured result, retrying a transient storage failure a bounded number of times. One locked or full write must not turn a finished answer into an ``unknown`` turn whose result lives only in memory. Retrying is exact: a write that committed before reporting failure is read back by ``finish_result``, and a failure after the commit (``settled`` set) never retries. A fence refusal (``RuntimeStoreError``) is not transient and is raised at once.r*NrZzFSettlement of admission %s deferred by storage (%r); retrying in %.1fsr ) r,r+_SETTLE_RETRY_DELAYS_Sr-r/r.rr0r1r2rerHr3) rr(rirr[rr\rr+delayexcs &&&&&&&& rsettle_with_retryrs?0)040 ' y'BSBS"ISxRZCh)ijj j1j' '} Z 7 NNc~. <--& & &  'sMCC?C C CCc VPVP,pVPP;_uu_4\ WWW4V4RRR4\ W4VP V4R# +'giL.;i)zKCleanup cannot withhold a committed outcome; publication is attempted once.N)rrcr&r'rkr_publish_pending)rrirrZr[r\r(s&&&&&& rpublish_terminalrsZ   cnn -D     ISw(S 9" s# s A11 B )N)g?g@)__doc__r-loggingr1hermes_state_runtimerrrr getLogger__name__rer"rr5r_rkrsrtrgr{rrrrrrrrs}S^^   8 $.*W #0f 8 NL*C`#F$'($r