vh
ff1f9ca1e8
- #368 (user,character) memory silo test DONE + PASSED live (WT b127): write-side conjunctive {end_user,agent_self} scoping + read-side cross- character isolation both proven end-to-end; betty (throwaway) deleted, Sindra intact. Full record in persistent-memory.d/2026-07-18-368-silo- test-passed.md. Retired the stale "silo test in progress" in-flight blocks. - Two-tier migration: split 152 over-threshold dated entries into persistent-memory.d/ detail files, leaving one-line pointers in the index (startup load ~196KB -> ~53KB; bodies now load on demand). - Tier-3 stores scrubbed clean (memory 0 / affect 0, provider restarted empty); persistent-memory + detail file updated to reflect the scrub.
1.3 KiB
[2026-07-10] Reverting our provider to bifrost 1.0.0 to fix the bound-handshake schema_validation_failed — DISPROVEN, and it would've been the WRONG state. Hypothesis: "my 1.1.0 bump broke the handshake; revert fixes it." Reverted 1.1.0→1.0.0, restarted :8392, retested → STILL failed (1.0.0 fails too). Root cause was Worldtree-side (their bifrost 0.9→1.1 bump in b35 broke frozen-v0.6), fixed by WT adopting wire-v0.7 (b47); staying on 1.1.0/v0.7 was the RIGHT state (1.0.0/v0.6 would've been incompatible with b47). Lessons: (1) don't bump a WIRE-PROTOCOL dependency out-of-lockstep with the peer on the other end of the wire; (2) a library's "additive/non-breaking over frozen vN" claim can fail if it emits new fields regardless of the negotiated version — bifrost 1.1.0 emitted sortable_chunk_fields on a v0.6 handshake; (3) diagnosis foot-gun: our provider logged handshake 200 (WE accepted) but WT rejected our RESPONSE, so the error surfaces at session-create direction-ambiguous — op-feed the handshake req/resp to disambiguate our-provider-rejects-request (i) vs peer-rejects-our-response (ii). Also: this sandbox BLOCKS foreground sleep (SIGTERMs the command, exit 144) — use separate tool calls / Monitor-until-loop for waits, never sleep in a compound command.**