@karmaniverous/jeeves-watcher-openclaw 0.4.1 → 0.4.2

package/dist/skills/jeeves-watcher/SKILL.md

@@ -96,28 +96,6 @@ You have two complementary tools with different scopes:
 
 **The principle:** Memory-core is your curated highlights. The watcher archive is your perfect recall. Use memory first for speed and signal, but never let its narrow scope be the ceiling of what you can remember.
 
-## Embedding Alignment
-
-Memory-core and the watcher both use embeddings, but may use different models by default. For best cross-tool consistency, offer to configure memory-core to use `gemini-embedding-001` (the same model the watcher uses) via the gateway config:
-
-1. **Verify Google API key** — check that the gateway has a Google API key configured (needed for Gemini embeddings).
-2. **Apply config** — use `gateway config.patch` to set the memory-core embedding provider:
-   ```json
-   {
-     "agents": {
-       "defaults": {
-         "memorySearch": {
-           "provider": "gemini",
-           "model": "gemini-embedding-001"
-         }
-       }
-     }
-   }
-   ```
-3. **Restart gateway** to pick up the new embedding provider. Memory-core will re-embed all memory files on next sync (dimension change triggers automatic vector table recreation).
-
-This gives memory-core the same 3072-dimensional Gemini embeddings the watcher uses, ensuring semantic similarity scores are comparable across memory and archive searches.
-
 ## Plugin Installation
 
 ```
@@ -165,14 +143,27 @@ If not running, install it. **Prefer native installation** (especially on cloud
 
 **Linux (recommended for servers):**
 ```bash
-# Download latest release
-curl -L https://github.com/qdrant/qdrant/releases/latest/download/qdrant-x86_64-unknown-linux-musl.tar.gz -o qdrant.tar.gz
-tar xzf qdrant.tar.gz
-
-# Run (foreground for testing)
-./qdrant
+# Download and install binary
+curl -L https://github.com/qdrant/qdrant/releases/latest/download/qdrant-x86_64-unknown-linux-musl.tar.gz -o /tmp/qdrant.tar.gz
+sudo tar xzf /tmp/qdrant.tar.gz -C /usr/local/bin/
+
+# Create qdrant user and directories
+sudo useradd -r -s /bin/false qdrant
+sudo mkdir -p /var/lib/qdrant/storage /var/lib/qdrant/snapshots /etc/qdrant
+sudo chown -R qdrant:qdrant /var/lib/qdrant
+
+# Create config
+sudo tee /etc/qdrant/config.yaml > /dev/null <<EOF
+storage:
+  storage_path: /var/lib/qdrant/storage
+  snapshots_path: /var/lib/qdrant/snapshots
+service:
+  host: 0.0.0.0
+  http_port: 6333
+  grpc_port: 6334
+EOF
 
-# For production: create a systemd service
+# Create systemd service
 sudo tee /etc/systemd/system/qdrant.service > /dev/null <<EOF
 [Unit]
 Description=Qdrant Vector Database
@@ -181,12 +172,14 @@ After=network.target
 [Service]
 Type=simple
 ExecStart=/usr/local/bin/qdrant --config-path /etc/qdrant/config.yaml
+WorkingDirectory=/var/lib/qdrant
 Restart=always
 User=qdrant
 
 [Install]
 WantedBy=multi-user.target
 EOF
+sudo systemctl daemon-reload
 sudo systemctl enable --now qdrant
 ```
 
@@ -310,6 +303,7 @@ After=network.target qdrant.service
 [Service]
 Type=simple
 ExecStart=$(which jeeves-watcher) start -c <config-path>
+WorkingDirectory=%h
 Restart=always
 Environment=GOOGLE_API_KEY=<key>
 User=$USER
@@ -351,10 +345,6 @@ Once health is confirmed and initial indexing has started:
 2. Query `$.inferenceRules[*].['name','description']` for available record types.
 3. Report to the user: how many points indexed so far, which domains are available, estimated time to complete initial indexing (based on file count and embedding rate).
 
-### Step 9: Align Memory-Core Embeddings
-
-After the watcher is healthy, offer to align OpenClaw's memory-core with the same embedding model for consistent vector quality across both systems. See the **Embedding Alignment** section above for the procedure.
-
 ### On Subsequent Sessions
 
 On sessions after bootstrap is complete:

package/openclaw.plugin.json

@@ -2,7 +2,7 @@
   "id": "jeeves-watcher-openclaw",
   "name": "Jeeves Watcher",
   "description": "Semantic search, metadata enrichment, and instance administration for a jeeves-watcher deployment.",
-  "version": "0.4.1",
+  "version": "0.4.2",
   "skills": [
     "dist/skills/jeeves-watcher"
   ],

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@karmaniverous/jeeves-watcher-openclaw",
-  "version": "0.4.1",
+  "version": "0.4.2",
   "author": "Jason Williscroft",
   "description": "OpenClaw plugin for jeeves-watcher — semantic search and metadata enrichment tools",
   "license": "BSD-3-Clause",