You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Demonstrate mixed control using executed mappings and backend evidence #1018
Produce a reproducible demonstration of the repaired shared-runtime control and mixed-execution paths against explicitly selected backend realizations.
Acceptance criteria
Pin a supported capability matrix and executing backend artifacts; add native blockers for any unfinished backend realization they require. Include simultaneous mixed execution and a staged handoff; separate single-backend runs do not establish mixed support.
Exercise repeated control changes, a governed effect, a non-identity mapping, and time/order coordination. Include denial, refusal, stale handoff, and missing/failed delivery.
Compare runtime outcomes with backend realization/readback and show that prohibited effects did not occur. Preserve participant, episode, operation, mapping, timing, policy, and configuration identity.
Publish the runnable scenario, commands, evidence, and independent reproduction record for the selected configurations. Do not claim universal backend equivalence or simulation-to-operational transfer.
changed the title [-]Demonstrate mixed cross-backend participant control[/-][+]Demonstrate mixed control using executed mappings and backend evidence[/+]on Sep 21, 2026
Activity
[-]Demonstrate mixed cross-backend participant control[/-][+]Deferred: Demonstrate mixed cross-backend participant control[/+][-]Deferred: Demonstrate mixed cross-backend participant control[/-][+]Demonstrate mixed cross-backend participant control[/+][-]Demonstrate mixed cross-backend participant control[/-][+]Demonstrate mixed control using executed mappings and backend evidence[/+]