test Supermemory with LiteLLM Gemini models

servers/unraid/supermemory-poc/README.md

@@ -10,10 +10,12 @@ production Hindsight service or change the default Hermes profile.
   cache live under `/mnt/user/appdata/supermemory-poc/data`.
 - A dedicated address on Docker `br0`; port `6767` is not published on the
   unRaid host address.
-- Local multilingual `Xenova/bge-m3` embeddings at 1024 dimensions.
-- One embedding worker and ingest concurrency 1. This avoids treating known
-  concurrent-local-embedding instability as a retrieval-quality result.
-- Extraction/summarization uses an operator-supplied OpenAI-compatible LLM.
+- Remote OpenAI-compatible embeddings through HK LiteLLM using
+  `google/gemini-embedding-2` at 1024 dimensions.
+- Extraction, summarization, and other intelligent steps use
+  `gemini-3.5-flash-lite` through the same LiteLLM endpoint.
+- Ingest concurrency remains 1 so the first POC emphasizes correctness and
+  observable cost over throughput.
 - Resource ceiling: 4 CPUs and 8 GiB RAM. Supermemory Local keeps its corpus
   in memory, so RSS must be watched as the sample grows.
 
@@ -24,8 +26,10 @@ directory and re-ingest when comparing a different embedding plan.
 
 1. Confirm the chosen `SUPERMEMORY_IPV4_ADDRESS` is absent from both Arcane's
    `br0` attachments and the LAN neighbor/DHCP tables.
-2. Add `SUPERMEMORY_IPV4_ADDRESS`, `OPENAI_API_KEY`, and optional model/base URL
-   overrides to this Project's Arcane environment. Do not put secrets in Git,
+2. Add `SUPERMEMORY_IPV4_ADDRESS`, `OPENAI_API_KEY`,
+   `OPENAI_BASE_URL=https://api.opfo.xyz/v1`, and
+   `OPENAI_MODEL=gemini-3.5-flash-lite` to this Project's Arcane environment.
+   `OPENAI_API_KEY` is the HK LiteLLM bearer key. Do not put it in Git,
    Compose, activity notes, or chat.
 3. Confirm `/mnt/user/appdata` has room for the 298 MiB server binary, the
    multilingual model cache, data, and rollback copy.
@@ -45,7 +49,7 @@ image build is required.
 
 The server creates its client bearer token at:
 
-`/mnt/user/appdata/supermemory-poc/data/api-key`
+`/mnt/user/appdata/supermemory-poc/data-google-gemini-embedding-2-1024/api-key`
 
 Read it through an authorized unRaid/Arcane console without printing it into
 logs. Save it only as `SUPERMEMORY_API_KEY` in the `supermemory-lab` Hermes
@@ -77,17 +81,16 @@ Run `poc.py smoke` before importing any sampled production material. It checks
 the v3 document path plus the v4 search, profile, and conversation endpoints
 used by Hermes.
 
-### Current local-embedding blocker
+### Embedding plan and smoke gate
 
-On the unRaid POC host, `server-v0.0.8` remains healthy but its local embedding
-worker returns `unknown certificate verification error` when first loading
-`Xenova/bge-m3`. Supplying a Mozilla CA bundle through `SSL_CERT_FILE`,
-`SSL_CERT_DIR`, `NODE_EXTRA_CA_CERTS`, and Debian's standard CA path did not
-change the error; TLS verification was never disabled. Do not import sampled
-Hindsight or document data until either a remote OpenAI-compatible embedding
-endpoint or a separately approved local embedding sidecar passes the smoke
-test. The server also returns HTTP 409 when deleting documents stuck in the
-`embedding` state, so failed smoke rows may require a fresh POC data directory.
+The earlier local `Xenova/bge-m3` experiment is preserved under the old `data`
+directory because embedding plans cannot be changed in place. This deployment
+uses the fresh `data-google-gemini-embedding-2-1024` directory. The one-shot
+`supermemory-smoke` service verifies the locked embedding plan, performs a
+Chinese hybrid-search round trip, checks provenance metadata, and exercises
+the profile and conversation endpoints. It writes a sentinel after success so
+ordinary redeploys do not add duplicate smoke documents. Do not import sampled
+Hindsight or document material until this service exits successfully.
 
 ## Evaluation guardrails