Analyst synthesis · taxonomy
Runtime trust primitives arrange continuity—not a feature catalogue
Seven coherence primitives describe how agents earn reviewer dialogue across time: continuity field, portable evidence plane, verifier-safe exchange lane, interoperability envelope, governable autonomy boundary, continuity memory substrate, and survivability rehearsal overlay.
Structured illustration only—Ω.2 / Trust Lock / LP-004–LP-010 echoed as doctrine labels without expanding SDK semantics; carriesExecutionAuthority false in verifier-shaped examples.
{ "carriesExecutionAuthority": false }Primitive lattice (SVG)
Illustrative graph only—node labels live in the cards below; not a live topology or metrics view.
Runtime continuity field
Bounds how execution slices inherit digest-qualified timelines—topology continuity echoes evolve as declarative overlays, not autonomous graph mutation claims.
Portable evidence plane
Bundles nest lineage hashes forward so envelopes travel beside human workflows—Replay Studio posture illustrates compare intervals without importing protected payloads.
Verifier-safe exchange lane
Exchange artifacts remain paste/parse local—verification receipts summarize posture without asserting wire authority against external verifier networks.
Interoperability envelope
LP-010-shaped stitching declares how continuity fragments reference one another across systems—storytelling envelopes, not silent cross-control-plane coupling.
Governable autonomy boundary
Trust Lock doctrine separates observation/replay/verify modes from deploy/mutate verbs—authority stays explicit in structured guards.
Continuity memory substrate
Longitudinal attestations and persistence vocabulary qualify how reviewers reopen admissibility conversations—memory routes remain illustrative prose.
Survivability rehearsal overlay
Recovery sequencing references bounded replay divergence as rehearsal inputs—coherence-evolution-topology echoes describe posture arcs, not incident automation SLAs.