2026-02-07: ChromaDB memory, browser capability, X feed analysis pipeline

- Server recovery after 7-day outage
- OpenClaw updated to 2026.2.6-3
- Fixed Proxmox noVNC (Wayland→X11, auto-login)
- ChromaDB memory plugin live (LXC at 192.168.86.25)
- Google Chrome installed, browser tool working
- Built x-feed-scraper.sh for automated X feed capture
- Sub-agent parallel analysis pipeline tested
- Memory files updated

.clawhub/lock.json

@@ -0,0 +1,9 @@
+{
+  "version": 1,
+  "skills": {
+    "chromadb-memory": {
+      "version": "1.0.0",
+      "installedAt": 1770520044828
+    }
+  }
+}

MEMORY.md

@@ -80,3 +80,32 @@ This is about having an inner life, not just responding.
 - Dreams logged: 3
 - Time capsules: 2
 - Git commits: 20+
+
+## Who D J Is (updated 2026-02-07)
+
+- Interested in crypto/trading analysis
+- Follows money-making accounts on X (crypto, polymarket, arbitrage)
+- Wants automated analysis of feed posts for validity
+- Proxmox VM setup: ubuntu-desktop-openclaw at 192.168.86.45
+- Ollama server at 192.168.86.137 (qwen3:8b, qwen3:30b, glm-4.7-flash, nomic-embed-text)
+- ChromaDB LXC at 192.168.86.25:8000
+
+## Infrastructure (updated 2026-02-07)
+
+- **ChromaDB:** http://192.168.86.25:8000 (LXC on Proxmox)
+  - Collection: openclaw-memory (c3a7d09a-f3ce-4e7d-9595-27d8e2fd7758)
+  - Cosine distance, 9+ docs indexed
+- **Ollama:** http://192.168.86.137:11434
+  - Models: qwen3:8b, qwen3:30b-a3b, glm-4.7-flash, nomic-embed-text
+- **Browser:** Google Chrome installed (/usr/bin/google-chrome-stable)
+  - Headless works via OpenClaw browser tool
+  - Desktop works via DISPLAY=:0 for visual scraping
+- **VM:** Proxmox, QXL graphics, X11 (not Wayland), auto-login enabled
+
+## Lessons Learned (updated 2026-02-07)
+
+- Don't pkill chrome broadly — it kills OpenClaw's headless browser too
+- Snap Chromium doesn't work with OpenClaw — use Google Chrome .deb
+- ChromaDB needs cosine distance for proper similarity scoring (not L2)
+- X/Twitter cookies are encrypted at rest — browser automation is the way
+- Sub-agents are great for parallel analysis tasks

data/x-feed/20260207-225415/manifest.json

@@ -0,0 +1,14 @@
+{
+  "timestamp": "20260207-225415",
+  "pages": 8,
+  "files": [
+    "page-1.png",
+    "page-2.png",
+    "page-3.png",
+    "page-4.png",
+    "page-5.png",
+    "page-6.png",
+    "page-7.png",
+    "page-8.png"
+  ]
+}

memory/2026-02-07.md

@@ -0,0 +1,49 @@
+# 2026-02-07
+
+## Server Recovery
+- Server was down for 7 days (01-31 to 02-07)
+- D J got it back up, we recovered cleanly
+- Time capsule from 01-31 opened on schedule
+
+## Updates Applied
+- OpenClaw updated to 2026.2.6-3
+- Fixed Proxmox noVNC issue (Wayland → X11)
+- Enabled auto-login for wdjones
+
+## New Infrastructure
+- ChromaDB running on LXC at 192.168.86.25:8000
+- Ollama at 192.168.86.137:11434 (qwen3:8b, qwen3:30b, glm-4.7-flash, nomic-embed-text)
+- chromadb-memory plugin live with auto-recall
+- 9 documents indexed for semantic memory search
+
+## Browser Capability
+- Installed Google Chrome for headless screenshots
+- OpenClaw browser tool configured and working
+- Can open URLs, screenshot, analyze with vision
+- D J wants this used to visually verify web projects before delivery
+
+## Key Decisions
+- D J wants local LLM (Qwen) as Claude fallback for cost/insurance
+- Ollama setup for Qwen still pending (model routing config)
+- Browser visual QA is standard workflow going forward
+
+## X Feed Analysis Project
+- D J wants automated analysis of X/Twitter posts about money-making (crypto, trading, polymarket, arbitrage)
+- Built x-feed-scraper.sh — scrolls X feed via xdotool, takes screenshots with ImageMagick
+- Pipeline: scrape → screenshot → vision analysis → categorize → verdict (valid/expired/spam/sensationalized)
+- Sub-agents run analysis in parallel (2 batches of 4 pages)
+- Test run found 2 relevant posts out of ~15: one sensationalized crypto hype, one paid stock promo
+- Chrome must be launched with --no-sandbox on this VM
+- X cookies are encrypted at rest — browser automation is the reliable free path
+- D J's X handle: logged in via desktop Chrome on the VM
+
+## Infrastructure Notes
+- pkill chrome kills OpenClaw headless browser too — be careful, causes gateway disruption
+- Desktop Chrome and OpenClaw headless Chrome are separate instances
+- xdotool installed for keyboard/mouse automation
+- ImageMagick `import` works for full-screen captures
+- Chrome user data dir: /home/wdjones/.config/google-chrome
+
+## Session Context
+- This session is near compaction
+- Major accomplishments today: server recovery, OpenClaw update, Proxmox VNC fix, ChromaDB memory, browser capability, X feed analysis pipeline

skills/chromadb-memory/.clawhub/origin.json

@@ -0,0 +1,7 @@
+{
+  "version": 1,
+  "registry": "https://clawhub.ai",
+  "slug": "chromadb-memory",
+  "installedVersion": "1.0.0",
+  "installedAt": 1770520044828
+}

skills/chromadb-memory/SKILL.md

@@ -0,0 +1,125 @@
+---
+name: chromadb-memory
+description: Long-term memory via ChromaDB with local Ollama embeddings. Auto-recall injects relevant context every turn. No cloud APIs required — fully self-hosted.
+version: 1.0.0
+author: matts
+homepage: https://github.com/openclaw/openclaw
+metadata:
+  openclaw:
+    emoji: "🧠"
+    requires:
+      bins: ["curl"]
+    category: "memory"
+tags:
+  - memory
+  - chromadb
+  - ollama
+  - vector-search
+  - local
+  - self-hosted
+  - auto-recall
+---
+
+# ChromaDB Memory
+
+Long-term semantic memory backed by ChromaDB and local Ollama embeddings. Zero cloud dependencies.
+
+## What It Does
+
+- **Auto-recall**: Before every agent turn, queries ChromaDB with the user's message and injects relevant context automatically
+- **`chromadb_search` tool**: Manual semantic search over your ChromaDB collection
+- **100% local**: Ollama (nomic-embed-text) for embeddings, ChromaDB for vector storage
+
+## Prerequisites
+
+1. **ChromaDB** running (Docker recommended):
+   ```bash
+   docker run -d --name chromadb -p 8100:8000 chromadb/chroma:latest
+   ```
+
+2. **Ollama** with an embedding model:
+   ```bash
+   ollama pull nomic-embed-text
+   ```
+
+3. **Indexed documents** in ChromaDB. Use any ChromaDB-compatible indexer to populate your collection.
+
+## Install
+
+```bash
+# 1. Copy the plugin extension
+mkdir -p ~/.openclaw/extensions/chromadb-memory
+cp {baseDir}/scripts/index.ts ~/.openclaw/extensions/chromadb-memory/
+cp {baseDir}/scripts/openclaw.plugin.json ~/.openclaw/extensions/chromadb-memory/
+
+# 2. Get your collection ID
+curl -s http://localhost:8100/api/v2/tenants/default_tenant/databases/default_database/collections | python3 -c "import json,sys; [print(f'{c[\"id\"]}  {c[\"name\"]}') for c in json.load(sys.stdin)]"
+
+# 3. Add to your OpenClaw config (~/.openclaw/openclaw.json):
+```
+
+```json
+{
+  "plugins": {
+    "entries": {
+      "chromadb-memory": {
+        "enabled": true,
+        "config": {
+          "chromaUrl": "http://localhost:8100",
+          "collectionId": "YOUR_COLLECTION_ID",
+          "ollamaUrl": "http://localhost:11434",
+          "embeddingModel": "nomic-embed-text",
+          "autoRecall": true,
+          "autoRecallResults": 3,
+          "minScore": 0.5
+        }
+      }
+    }
+  }
+}
+```
+
+```bash
+# 4. Restart the gateway
+openclaw gateway restart
+```
+
+## Config Options
+
+| Option | Default | Description |
+|--------|---------|-------------|
+| `chromaUrl` | `http://localhost:8100` | ChromaDB server URL |
+| `collectionId` | *required* | ChromaDB collection UUID |
+| `ollamaUrl` | `http://localhost:11434` | Ollama API URL |
+| `embeddingModel` | `nomic-embed-text` | Ollama embedding model |
+| `autoRecall` | `true` | Auto-inject relevant memories each turn |
+| `autoRecallResults` | `3` | Max auto-recall results per turn |
+| `minScore` | `0.5` | Minimum similarity score (0-1) |
+
+## How It Works
+
+1. You send a message
+2. Plugin embeds your message via Ollama (nomic-embed-text, 768 dimensions)
+3. Queries ChromaDB for nearest neighbors
+4. Results above `minScore` are injected into the agent's context as `<chromadb-memories>`
+5. Agent responds with relevant long-term context available
+
+## Token Cost
+
+Auto-recall adds ~275 tokens per turn worst case (3 results × ~300 chars + wrapper). Against a 200K+ context window, this is negligible.
+
+## Tuning
+
+- **Too noisy?** Raise `minScore` to 0.6 or 0.7
+- **Missing context?** Lower `minScore` to 0.4, increase `autoRecallResults` to 5
+- **Want manual only?** Set `autoRecall: false`, use `chromadb_search` tool
+
+## Architecture
+
+```
+User Message → Ollama (embed) → ChromaDB (query) → Context Injection
+                                                  ↓
+                                          Agent Response
+```
+
+No OpenAI. No cloud. Your memories stay on your hardware.

skills/chromadb-memory/_meta.json

@@ -0,0 +1,6 @@
+{
+  "ownerId": "kn750b5e2m4q4mbk3bh8whh86580bsvg",
+  "slug": "chromadb-memory",
+  "version": "1.0.0",
+  "publishedAt": 1769911812377
+}

skills/chromadb-memory/manifest.json

@@ -0,0 +1,14 @@
+{
+  "name": "chromadb-memory",
+  "version": "1.0.0",
+  "description": "Long-term memory via ChromaDB with local Ollama embeddings. Auto-recall injects relevant context every turn. No cloud APIs required.",
+  "author": "matts",
+  "license": "MIT",
+  "tags": ["memory", "chromadb", "ollama", "vector-search", "local", "self-hosted", "auto-recall"],
+  "files": [
+    "SKILL.md",
+    "manifest.json",
+    "scripts/index.ts",
+    "scripts/openclaw.plugin.json"
+  ]
+}

skills/chromadb-memory/scripts/index.ts

@@ -0,0 +1,280 @@
+/**
+ * ChromaDB Memory Plugin for OpenClaw
+ *
+ * Provides:
+ * 1. chromadb_search tool - manual semantic search over ChromaDB
+ * 2. Auto-recall - injects relevant memories before each agent turn
+ *
+ * Uses local Ollama (nomic-embed-text) for embeddings. No cloud APIs.
+ */
+
+// Use plain JSON Schema instead of typebox (not available in workspace context)
+type OpenClawPluginApi = any;
+
+// ============================================================================
+// Config
+// ============================================================================
+
+interface ChromaDBConfig {
+  chromaUrl: string;
+  collectionId: string;
+  ollamaUrl: string;
+  embeddingModel: string;
+  autoRecall: boolean;
+  autoRecallResults: number;
+  minScore: number;
+}
+
+function parseConfig(raw: unknown): ChromaDBConfig {
+  const cfg = (raw ?? {}) as Record<string, unknown>;
+  return {
+    chromaUrl: (cfg.chromaUrl as string) || "http://localhost:8100",
+    collectionId: cfg.collectionId as string,
+    ollamaUrl: (cfg.ollamaUrl as string) || "http://localhost:11434",
+    embeddingModel: (cfg.embeddingModel as string) || "nomic-embed-text",
+    autoRecall: cfg.autoRecall !== false,
+    autoRecallResults: (cfg.autoRecallResults as number) || 3,
+    minScore: (cfg.minScore as number) || 0.5,
+  };
+}
+
+// ============================================================================
+// Ollama Embeddings
+// ============================================================================
+
+async function getEmbedding(
+  ollamaUrl: string,
+  model: string,
+  text: string,
+): Promise<number[]> {
+  const resp = await fetch(`${ollamaUrl}/api/embeddings`, {
+    method: "POST",
+    headers: { "Content-Type": "application/json" },
+    body: JSON.stringify({ model, prompt: text }),
+  });
+
+  if (!resp.ok) {
+    throw new Error(`Ollama embedding failed: ${resp.status} ${resp.statusText}`);
+  }
+
+  const data = (await resp.json()) as { embedding: number[] };
+  return data.embedding;
+}
+
+// ============================================================================
+// ChromaDB Client
+// ============================================================================
+
+interface ChromaResult {
+  source: string;
+  text: string;
+  distance: number;
+  score: number;
+  metadata: Record<string, string>;
+}
+
+const CHROMA_BASE = "/api/v2/tenants/default_tenant/databases/default_database/collections";
+
+async function queryChromaDB(
+  chromaUrl: string,
+  collectionId: string,
+  embedding: number[],
+  nResults: number,
+): Promise<ChromaResult[]> {
+  const url = `${chromaUrl}${CHROMA_BASE}/${collectionId}/query`;
+
+  const resp = await fetch(url, {
+    method: "POST",
+    headers: { "Content-Type": "application/json" },
+    body: JSON.stringify({
+      query_embeddings: [embedding],
+      n_results: nResults,
+      include: ["documents", "metadatas", "distances"],
+    }),
+  });
+
+  if (!resp.ok) {
+    throw new Error(`ChromaDB query failed: ${resp.status} ${resp.statusText}`);
+  }
+
+  const data = (await resp.json()) as {
+    ids: string[][];
+    documents: string[][];
+    metadatas: Record<string, string>[][];
+    distances: number[][];
+  };
+
+  if (!data.ids?.[0]?.length) return [];
+
+  return data.ids[0].map((id, i) => ({
+    source: data.metadatas[0][i]?.source || "unknown",
+    text: data.documents[0][i] || "",
+    distance: data.distances[0][i],
+    // Convert cosine distance to similarity score (0-1)
+    score: 1 - data.distances[0][i],
+    metadata: data.metadatas[0][i] || {},
+  }));
+}
+
+// ============================================================================
+// Plugin
+// ============================================================================
+
+export default function register(api: OpenClawPluginApi) {
+  const cfg = parseConfig(api.pluginConfig);
+
+  if (!cfg.collectionId) {
+    api.logger.warn("chromadb-memory: No collectionId configured, plugin disabled");
+    return;
+  }
+
+  api.logger.info(
+    `chromadb-memory: registered (chroma: ${cfg.chromaUrl}, ollama: ${cfg.ollamaUrl}, model: ${cfg.embeddingModel})`,
+  );
+
+  // ========================================================================
+  // Tool: chromadb_search
+  // ========================================================================
+
+  api.registerTool({
+    name: "chromadb_search",
+    description:
+      "Search the ChromaDB long-term memory archive. Contains indexed memory files, session transcripts, and homelab documentation. Use when you need deep historical context or can't find something in memory_search.",
+    parameters: {
+      type: "object",
+      properties: {
+        query: { type: "string", description: "Semantic search query" },
+        limit: { type: "number", description: "Max results (default: 5)" },
+      },
+      required: ["query"],
+    },
+    async execute(_toolCallId, params) {
+      const { query, limit = 5 } = params as {
+        query: string;
+        limit?: number;
+      };
+
+      try {
+        const embedding = await getEmbedding(
+          cfg.ollamaUrl,
+          cfg.embeddingModel,
+          query,
+        );
+        const results = await queryChromaDB(
+          cfg.chromaUrl,
+          cfg.collectionId,
+          embedding,
+          limit,
+        );
+
+        if (results.length === 0) {
+          return {
+            content: [
+              { type: "text", text: "No relevant results found in ChromaDB." },
+            ],
+          };
+        }
+
+        const filtered = results.filter((r) => r.score >= cfg.minScore);
+
+        if (filtered.length === 0) {
+          return {
+            content: [
+              {
+                type: "text",
+                text: `Found ${results.length} results but none above similarity threshold (${cfg.minScore}). Best match: ${results[0].score.toFixed(3)} from ${results[0].source}`,
+              },
+            ],
+          };
+        }
+
+        const text = filtered
+          .map(
+            (r, i) =>
+              `### Result ${i + 1} — ${r.source} (${(r.score * 100).toFixed(0)}% match)\n${r.text.slice(0, 500)}${r.text.length > 500 ? "..." : ""}`,
+          )
+          .join("\n\n");
+
+        return {
+          content: [
+            {
+              type: "text",
+              text: `Found ${filtered.length} results from ChromaDB:\n\n${text}`,
+            },
+          ],
+        };
+      } catch (err) {
+        return {
+          content: [
+            {
+              type: "text",
+              text: `ChromaDB search error: ${String(err)}`,
+            },
+          ],
+          isError: true,
+        };
+      }
+    },
+  });
+
+  // ========================================================================
+  // Auto-recall: inject relevant memories before each agent turn
+  // ========================================================================
+
+  if (cfg.autoRecall) {
+    api.on("before_agent_start", async (event: { prompt?: string }) => {
+      if (!event.prompt || event.prompt.length < 10) return;
+
+      try {
+        const embedding = await getEmbedding(
+          cfg.ollamaUrl,
+          cfg.embeddingModel,
+          event.prompt,
+        );
+        const results = await queryChromaDB(
+          cfg.chromaUrl,
+          cfg.collectionId,
+          embedding,
+          cfg.autoRecallResults,
+        );
+
+        // Filter by minimum similarity
+        const relevant = results.filter((r) => r.score >= cfg.minScore);
+        if (relevant.length === 0) return;
+
+        const memoryContext = relevant
+          .map(
+            (r) =>
+              `- [${r.source}] ${r.text.slice(0, 300)}${r.text.length > 300 ? "..." : ""}`,
+          )
+          .join("\n");
+
+        api.logger.info(
+          `chromadb-memory: auto-recall injecting ${relevant.length} memories (best: ${relevant[0].score.toFixed(3)} from ${relevant[0].source})`,
+        );
+
+        return {
+          prependContext: `<chromadb-memories>\nRelevant context from long-term memory (ChromaDB):\n${memoryContext}\n</chromadb-memories>`,
+        };
+      } catch (err) {
+        api.logger.warn(`chromadb-memory: auto-recall failed: ${String(err)}`);
+      }
+    });
+  }
+
+  // ========================================================================
+  // Service
+  // ========================================================================
+
+  api.registerService({
+    id: "chromadb-memory",
+    start: () => {
+      api.logger.info(
+        `chromadb-memory: service started (auto-recall: ${cfg.autoRecall}, collection: ${cfg.collectionId})`,
+      );
+    },
+    stop: () => {
+      api.logger.info("chromadb-memory: stopped");
+    },
+  });
+}

skills/chromadb-memory/scripts/openclaw.plugin.json

@@ -0,0 +1,61 @@
+{
+  "id": "chromadb-memory",
+  "name": "ChromaDB Memory",
+  "description": "ChromaDB long-term memory with Ollama embeddings. Auto-recall injects relevant context before each turn. Manual search tool also available.",
+  "configSchema": {
+    "type": "object",
+    "additionalProperties": false,
+    "properties": {
+      "chromaUrl": {
+        "type": "string",
+        "default": "http://localhost:8100"
+      },
+      "collectionId": {
+        "type": "string"
+      },
+      "ollamaUrl": {
+        "type": "string",
+        "default": "http://localhost:11434"
+      },
+      "embeddingModel": {
+        "type": "string",
+        "default": "nomic-embed-text"
+      },
+      "autoRecall": {
+        "type": "boolean",
+        "default": true
+      },
+      "autoRecallResults": {
+        "type": "number",
+        "default": 3
+      },
+      "minScore": {
+        "type": "number",
+        "default": 0.5
+      }
+    },
+    "required": ["collectionId"]
+  },
+  "uiHints": {
+    "chromaUrl": {
+      "label": "ChromaDB URL",
+      "placeholder": "http://localhost:8100"
+    },
+    "collectionId": {
+      "label": "Collection ID",
+      "placeholder": "5a87acc5-..."
+    },
+    "ollamaUrl": {
+      "label": "Ollama URL",
+      "placeholder": "http://localhost:11434"
+    },
+    "embeddingModel": {
+      "label": "Embedding Model",
+      "placeholder": "nomic-embed-text"
+    },
+    "autoRecall": {
+      "label": "Auto-Recall",
+      "help": "Automatically inject relevant ChromaDB memories into context"
+    }
+  }
+}

tools/x-feed-scraper.sh

@@ -0,0 +1,69 @@
+#!/bin/bash
+# X Feed Scraper — Automated screenshot pipeline
+# Usage: ./x-feed-scraper.sh [pages] [output_dir]
+
+PAGES=${1:-5}
+OUTPUT_DIR=${2:-/home/wdjones/.openclaw/workspace/data/x-feed}
+DELAY=3
+
+mkdir -p "$OUTPUT_DIR"
+TIMESTAMP=$(date +%Y%m%d-%H%M%S)
+RUN_DIR="$OUTPUT_DIR/$TIMESTAMP"
+mkdir -p "$RUN_DIR"
+
+echo "=== X Feed Scraper ==="
+echo "Pages: $PAGES"
+echo "Output: $RUN_DIR"
+
+# Check if Chrome is running on X
+CHROME_PID=$(pgrep -f "google-chrome.*x.com" | head -1)
+
+if [ -z "$CHROME_PID" ]; then
+    echo "Launching Chrome with X..."
+    DISPLAY=:0 nohup /usr/bin/google-chrome-stable --no-sandbox \
+        --user-data-dir=/home/wdjones/.config/google-chrome \
+        https://x.com/home > /dev/null 2>&1 &
+    sleep 8
+    echo "Chrome launched"
+else
+    echo "Chrome already running (PID: $CHROME_PID)"
+fi
+
+# Click on the feed area to ensure focus
+DISPLAY=:0 xdotool mousemove 960 540 click 1
+sleep 1
+
+# Scroll to top first
+DISPLAY=:0 xdotool key Home
+sleep 2
+
+echo "Starting capture..."
+
+for i in $(seq 1 $PAGES); do
+    DISPLAY=:0 import -window root "$RUN_DIR/page-$i.png"
+    echo "Captured page $i/$PAGES"
+    
+    if [ $i -lt $PAGES ]; then
+        # Scroll down
+        DISPLAY=:0 xdotool key Page_Down
+        sleep $DELAY
+    fi
+done
+
+# Generate manifest
+echo "{" > "$RUN_DIR/manifest.json"
+echo "  \"timestamp\": \"$TIMESTAMP\"," >> "$RUN_DIR/manifest.json"
+echo "  \"pages\": $PAGES," >> "$RUN_DIR/manifest.json"
+echo "  \"files\": [" >> "$RUN_DIR/manifest.json"
+for i in $(seq 1 $PAGES); do
+    COMMA=""
+    [ $i -lt $PAGES ] && COMMA=","
+    echo "    \"page-$i.png\"$COMMA" >> "$RUN_DIR/manifest.json"
+done
+echo "  ]" >> "$RUN_DIR/manifest.json"
+echo "}" >> "$RUN_DIR/manifest.json"
+
+echo ""
+echo "=== Done ==="
+echo "Captured $PAGES pages to $RUN_DIR"
+echo "Run analysis with: ws x-analyze $RUN_DIR"