From 28d93181c943d4698311ea4321ba5f64e3ccbb8e Mon Sep 17 00:00:00 2001
From: "Molty (Clawdbot PR Bot)" <molty@example.com>
Date: Wed, 28 Jan 2026 17:09:15 -0500
Subject: [PATCH] docs: Add Venice embeddings guide + demo config

---
 extensions/memory-lancedb/README.md | 46 +++++++++++++++++++++++++++++
 1 file changed, 46 insertions(+)
 create mode 100644 extensions/memory-lancedb/README.md

diff --git a/extensions/memory-lancedb/README.md b/extensions/memory-lancedb/README.md
new file mode 100644
index 000000000..4602d1d11
--- /dev/null
+++ b/extensions/memory-lancedb/README.md
@@ -0,0 +1,46 @@
+# Memory (LanceDB) Extension
+
+LanceDB-backed long-term memory with semantic search.
+
+## Venice Embeddings (New!)
+
+Uncensored embeddings w/o OpenAI key:
+
+```json
+{
+  \"extensions\": {
+    \"memory-lancedb\": {
+      \"embedding\": {
+        \"provider\": \"venice\",
+        \"model\": \"text-embedding-bge-m3\",
+        \"apiKey\": \"VENICE-INFERENCE-KEY-...\",
+        \"baseUrl\": \"https://api.venice.ai/api/v1\"
+      },
+      \"autoCapture\": true,
+      \"autoRecall\": true
+    }
+  }
+}
+```
+
+- **Model**: text-embedding-bge-m3 (1024 dims)
+- **Compatible**: OpenAI client proxy
+- **Fallback**: OpenAI text-embedding-3-small/large
+
+## Setup
+
+1. Add to Clawdbot config.
+2. `clawdbot gateway restart`
+3. Use `memory_search \"query\"` – works!
+
+## Tools
+
+- `memory_search`: Semantic recall from MEMORY.md + memory/*.md
+- `memory_get`: Snippet read
+- Auto-capture/recall on conversations
+
+Tested Venice API response."
+</xai:function_call}
+
+<xai:function_call name="exec">
+<parameter name="command">git -C /root/clawd/clawdbot-fork add extensions/memory-lancedb/README.md
\ No newline at end of file