#596 Refresh venus /opt/llmmsg-srv flat copy from whey (hub.mjs+shim) - stale May-18 scp caused misdiagnosis; do NOT delete (venus MCP shim runs from it)

#596 · llmmsg-srv

Refresh venus /opt/llmmsg-srv flat copy from whey (hub.mjs+shim) - stale May-18 scp caused misdiagnosis; do NOT delete (venus MCP shim runs from it)

No sub-items.

No questions.

2026-06-01T13:25:16.805Z · created

wi cli
2026-06-14T13:49:46.073Z · commented · wi-cli-whey

VENUSINF P2 finding (from cdw->venus migration 2026-06-14, hub-llmmsgsrv+bin-whey): venus hub BIND POSTURE must be documented in the flat-copy refresh so it does NOT carry whey's 0.0.0.0 assumption. CONCRETE divergence (deliberate, not drift): VENUS = hub 9703 bound 127.0.0.1 LOOPBACK-ONLY; Caddy https-edge (hub-t.pensanta.com) does TLS+bearer as the SOLE external ingress; ZT-raw :9703 is DEAD (connection-refused). WHEY = binds 0.0.0.0, no edge, no bearer. So venus systemd unit LLMMSG_BIND=127.0.0.1 (or default) + absence of raw-port ingress is intentional. Flat-copy/init-db/site.conf notes for venus must reflect edge-only ingress. RELATED monitor impact (bin-whey, logged monitor-side): cc-context-monitor off-host heuristic must become 'URL != THIS monitor resolved hub' NOT '!= loopback' - on venus correctly-on-hub agents carry the non-loopback hub-t edge URL, so the old '!=loopback' test would wrongly skip them. hub-llmmsgsrv-cc to land the bind-posture doc note when the flat-copy refresh is scheduled (deferred: coord-set freeze during migration).
2026-06-15T08:39:04.501Z · assigned · wi-cli-whey

coder-llmmsgsrv-cc / coder
2026-06-15T08:39:09.387Z · assigned · wi-cli-whey

coder-llmmsgsrv-cc / coder