memory: demo rolled back from broken c2d87263 to fa8bc51c; b193 retired-key list for demo

This commit is contained in:
vh
2026-10-02 08:17:24 -07:00
parent 8b81579bca
commit 1219cfaca9
2 changed files with 2 additions and 1 deletions
@@ -36,7 +36,7 @@
- `policies.yaml`: drop the memory-dream-allow rule and admin-everything's `exclude_actions: [memory.dream]` (the action is gone; a policy naming it is inert). - `policies.yaml`: drop the memory-dream-allow rule and admin-everything's `exclude_actions: [memory.dream]` (the action is gone; a policy naming it is inert).
- `model_roles.yaml`: drop `memory_distiller` AND `memory_extractor`. - `model_roles.yaml`: drop `memory_distiller` AND `memory_extractor`.
- `providers.yaml`: KEEP the `summarizer` catalog entry (the deprecated granite-4.1-8b preset still binds `catalog_id: summarizer`; dangling = boot error); only its description changes. - `providers.yaml`: KEEP the `summarizer` catalog entry (the deprecated granite-4.1-8b preset still binds `catalog_id: summarizer`; dangling = boot error); only its description changes.
- `defaults.yaml`: drop the stale `memory.legacy.mode` comment block, the WHOLE `memory.escalation` block, and any other key b193's boot WARNINGs name as retired (sweep both api logs after the push). - `defaults.yaml`: drop the stale `memory.legacy.mode` comment block and every key b193's boot WARNINGs name as retired. **Demo's list, read from its b193 boot 2026-10-02 15:16Z:** `memory.agent_self_enabled`, `memory.backend`, `memory.dream`, `memory.escalation`, `memory.legacy`, `memory.tier3_embedding_provider`, `memory.triggers`. Personal: read its boot log once it is on b193.
- **`scripts/wt-memory-gate-batch` must drop `--legacy-mode` when b193 goes live** (the flag no longer exists, so infra-hermes' daily batch would fail; worldtree-dev msg 6653). Change it in the same window as the b193 deploy. - **`scripts/wt-memory-gate-batch` must drop `--legacy-mode` when b193 goes live** (the flag no longer exists, so infra-hermes' daily batch would fail; worldtree-dev msg 6653). Change it in the same window as the b193 deploy.
- No action, for the record (msg 6380): at b193 `memory.writer.enabled` / `memory.reader.enabled` default to TRUE. Both instances set them `true` explicitly; KEEP those lines (explicit over implicit). A `false` would now boot with one WARNING. - No action, for the record (msg 6380): at b193 `memory.writer.enabled` / `memory.reader.enabled` default to TRUE. Both instances set them `true` explicitly; KEEP those lines (explicit over implicit). A `false` would now boot with one WARNING.
- Why `memory_extractor` can go: `memory.writer.seat` is unset on both (effective `'agent'`, evaluated with `WriterConfig` in both running apis). The other reference, `memory.escalation.model_role`, is inert at b193: `escalation` is in `RETIRED_MEMORY_KEYS`, never parsed, one boot WARNING (worldtree-dev msg 6314). memory_tagger is the only memory role left. - Why `memory_extractor` can go: `memory.writer.seat` is unset on both (effective `'agent'`, evaluated with `WriterConfig` in both running apis). The other reference, `memory.escalation.model_role`, is inert at b193: `escalation` is in `RETIRED_MEMORY_KEYS`, never parsed, one boot WARNING (worldtree-dev msg 6314). memory_tagger is the only memory role left.
+1
View File
@@ -287,6 +287,7 @@ _As of 2026-10-01 ~0446 PT._
## Recent decisions ## Recent decisions
- `[2026-10-02]` **Demo outage ~13 min, ROLLED BACK (worldtree-dev URGENT 6684).** Their b193 release commit c2d87263 swept 60 staged deletions into the tree, so the demo api crash-looped. I recreated `worldtree-api` only (`compose up -d --no-deps`) on the `.env`-pinned fa8bc51cc064 (compose.yaml identical at both shas): healthy in 35 s at 15:16Z. ⚠ Their deploy health gate does NOT restore the old container on failure (it only withholds `:latest`); flagged to them. Fix-forward comes through CI.
- `[2026-10-02]` **nh3-pve's AMT is in TacticalRMM's MeshCentral (group `PFI-AMT`, device `nh3-pve-amt`).** The adds had failed silently because TRMM installs MeshCentral `WANonly` (meshuser.js:2682 drops AMT adds). Prime ruled hybrid; switched 0812, 14/14 agents back, AMT connected at once (16.1.25, TLS fine). ⚠ Still reached via nh3-scale ON nh3-pve, so useless when nh3-pve is down; the fix is the IPsec route or CIRA (MPS :4433 is live). 2FA still not forced. Prime's MeshCentral login token is vaulted `pfi-tacticalrmm/meshcentral-login-token`. → `servers/pfi-tacticalrmm/README.md` - `[2026-10-02]` **nh3-pve's AMT is in TacticalRMM's MeshCentral (group `PFI-AMT`, device `nh3-pve-amt`).** The adds had failed silently because TRMM installs MeshCentral `WANonly` (meshuser.js:2682 drops AMT adds). Prime ruled hybrid; switched 0812, 14/14 agents back, AMT connected at once (16.1.25, TLS fine). ⚠ Still reached via nh3-scale ON nh3-pve, so useless when nh3-pve is down; the fix is the IPsec route or CIRA (MPS :4433 is live). 2FA still not forced. Prime's MeshCentral login token is vaulted `pfi-tacticalrmm/meshcentral-login-token`. → `servers/pfi-tacticalrmm/README.md`
- `[2026-10-02]` **Prime: dev-backup gets dailies + weeklies — DONE.** Retention is now 48 hourly + newest of 30 days + newest of 12 ISO weeks (`retention.py` beside the script; second path guard on the NAS side; unit fails if kept ≠ expected). Live run 0739: deleted 1, 0 errors, 48 kept = expected. History before 09-30 was already gone; dailies accumulate from today. - `[2026-10-02]` **Prime: dev-backup gets dailies + weeklies — DONE.** Retention is now 48 hourly + newest of 30 days + newest of 12 ISO weeks (`retention.py` beside the script; second path guard on the NAS side; unit fails if kept ≠ expected). Live run 0739: deleted 1, 0 errors, 48 kept = expected. History before 09-30 was already gone; dailies accumulate from today.
- `[2026-10-02]` **Prime: MS-03s get Proxmox; dummy plugs in hand.** Still open: what the second MS-03 is for. On arrival: check the NIC chipset (I226-LM = keep the AMT port admin-UP), fit a plug on each, then the parked AMT follow-ups. - `[2026-10-02]` **Prime: MS-03s get Proxmox; dummy plugs in hand.** Still open: what the second MS-03 is for. On arrival: check the NIC chipset (I226-LM = keep the AMT port admin-UP), fit a plug on each, then the parked AMT follow-ups.