Fix Artemis's own CD-INIT regression, same class as Hermes's

Found while auditing messaging capabilities: artemis/init.4th defines its own CD-INIT (block 4141, "Hera calls this at fleet boot") but nothing calls it -- same latent gap as Hermes's CD-INIT (fixed last session), just undiscovered because the earlier check only searched for ART-BOOT-ENTRY, not CD-INIT specifically. ART-BOOT-ENTRY runs unconditionally but never calls CD-INIT internally. Added an unconditional call in the same trailing block that already calls STARTUP-BANNER.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019ZGkimpfyh63EZyRkNbkPD

capsules/artemis/init.4th

@@ -488,6 +488,7 @@ Block 4852
 ( workload, for doe-campaign.4th's remote VM-EXEC. )
 : LOAD-DOE ( -- ) S" doe.4th" EXEC ;
 Block 4853
-( D.7: Artemis didn't load lib.4th before; needs it now. )
+( D.7: CD-INIT ran only via the old self-test; birth needs it. )
+CD-INIT
 S" lib.4th" EXEC
 STARTUP-BANNER