zubo 0.1.4 → 0.1.6

package/.dockerignore

@@ -0,0 +1,12 @@
+node_modules
+dist
+desktop
+site
+.env
+.env.*
+.zubo
+.git
+.DS_Store
+*.db
+*.db-wal
+*.db-shm

package/Dockerfile

@@ -0,0 +1,24 @@
+# Zubo — Dockerfile
+# Runs the Zubo AI agent using the official Bun runtime image.
+# Persistent data (config, database, memory, skills) lives in /root/.zubo
+# and should be mounted as a volume.
+
+FROM oven/bun:1
+
+WORKDIR /app
+
+# Install dependencies first (layer cache optimisation)
+COPY package.json bun.lock ./
+RUN bun install --frozen-lockfile
+
+# Copy the rest of the project
+COPY . .
+
+# Webchat dashboard port
+EXPOSE 8787
+
+# Persistent storage mount point — config.json, zubo.db, workspace, logs, etc.
+VOLUME /root/.zubo
+
+# Start the agent
+CMD ["bun", "run", "src/index.ts", "start"]

package/docker-compose.yml

@@ -0,0 +1,20 @@
+# Zubo — Docker Compose
+# Usage:
+#   1. Copy .env.example to .env and fill in your API keys
+#   2. docker compose up -d
+#   3. Open http://localhost:8787 for the webchat dashboard
+
+services:
+  zubo:
+    build: .
+    container_name: zubo
+    restart: unless-stopped
+
+    ports:
+      - "8787:8787"   # Webchat dashboard
+
+    volumes:
+      - ~/.zubo:/root/.zubo   # Persistent data (config, DB, memory, skills, logs)
+
+    env_file:
+      - .env   # API keys: ANTHROPIC_API_KEY, OPENAI_API_KEY, TELEGRAM_BOT_TOKEN, etc.

package/migrations/015_webhooks.sql

@@ -0,0 +1,19 @@
+CREATE TABLE IF NOT EXISTS webhooks (
+  id TEXT PRIMARY KEY,
+  name TEXT NOT NULL UNIQUE,
+  description TEXT,
+  secret TEXT,
+  active INTEGER NOT NULL DEFAULT 1,
+  created_at TEXT NOT NULL DEFAULT (datetime('now'))
+);
+
+CREATE TABLE IF NOT EXISTS webhook_events (
+  id INTEGER PRIMARY KEY AUTOINCREMENT,
+  webhook_id TEXT NOT NULL REFERENCES webhooks(id) ON DELETE CASCADE,
+  payload TEXT NOT NULL,
+  headers TEXT,
+  processed INTEGER NOT NULL DEFAULT 0,
+  created_at TEXT NOT NULL DEFAULT (datetime('now'))
+);
+
+CREATE INDEX IF NOT EXISTS idx_webhook_events_webhook ON webhook_events(webhook_id);

package/migrations/016_knowledge_graph.sql

@@ -0,0 +1,23 @@
+CREATE TABLE IF NOT EXISTS kg_entities (
+  id INTEGER PRIMARY KEY AUTOINCREMENT,
+  name TEXT NOT NULL,
+  type TEXT NOT NULL,
+  properties TEXT,
+  created_at TEXT NOT NULL DEFAULT (datetime('now')),
+  updated_at TEXT NOT NULL DEFAULT (datetime('now'))
+);
+
+CREATE UNIQUE INDEX IF NOT EXISTS idx_kg_entity_name_type ON kg_entities(name, type);
+
+CREATE TABLE IF NOT EXISTS kg_relations (
+  id INTEGER PRIMARY KEY AUTOINCREMENT,
+  source_id INTEGER NOT NULL REFERENCES kg_entities(id) ON DELETE CASCADE,
+  target_id INTEGER NOT NULL REFERENCES kg_entities(id) ON DELETE CASCADE,
+  relation TEXT NOT NULL,
+  properties TEXT,
+  created_at TEXT NOT NULL DEFAULT (datetime('now'))
+);
+
+CREATE INDEX IF NOT EXISTS idx_kg_rel_source ON kg_relations(source_id);
+CREATE INDEX IF NOT EXISTS idx_kg_rel_target ON kg_relations(target_id);
+CREATE INDEX IF NOT EXISTS idx_kg_rel_type ON kg_relations(relation);

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "zubo",
-  "version": "0.1.4",
-  "description": "Your AI agent that never forgets. Persistent memory, 20+ tools, 6 channels, 11+ LLM providers — runs entirely on your machine.",
+  "version": "0.1.6",
+  "description": "Your AI agent that never forgets. Persistent memory, 25+ tools, 7 channels, 11+ LLM providers — runs entirely on your machine.",
   "license": "MIT",
   "author": "thomaskanze",
   "homepage": "https://zubo.bot",

package/site/docs/index.html

@@ -4,18 +4,18 @@
   <meta charset="UTF-8">
   <meta name="viewport" content="width=device-width, initial-scale=1.0">
   <title>Getting Started — Zubo Docs</title>
-  <meta name="description" content="Get started with Zubo, a self-hosted AI agent with persistent semantic memory, multi-channel presence, 35+ tools, sub-agent delegation, and workflow automation.">
+  <meta name="description" content="Get started with Zubo, a self-hosted AI agent with persistent semantic memory, multi-channel presence, 25+ tools, sub-agent delegation, MCP support, and workflow automation.">
   <meta name="theme-color" content="#060608">
   <link rel="canonical" href="https://zubo.bot/docs/">
   <meta property="og:title" content="Getting Started — Zubo Docs">
-  <meta property="og:description" content="Get started with Zubo, a self-hosted AI agent with persistent semantic memory, multi-channel presence, 35+ tools, sub-agent delegation, and workflow automation.">
+  <meta property="og:description" content="Get started with Zubo, a self-hosted AI agent with persistent semantic memory, multi-channel presence, 25+ tools, sub-agent delegation, MCP support, and workflow automation.">
   <meta property="og:type" content="article">
   <meta property="og:url" content="https://zubo.bot/docs/">
   <meta property="og:image" content="https://zubo.bot/og-image.png">
   <meta property="og:site_name" content="Zubo">
   <meta name="twitter:card" content="summary_large_image">
   <meta name="twitter:title" content="Getting Started — Zubo Docs">
-  <meta name="twitter:description" content="Get started with Zubo, a self-hosted AI agent with persistent semantic memory, 35+ tools, and workflow automation.">
+  <meta name="twitter:description" content="Get started with Zubo, a self-hosted AI agent with persistent semantic memory, 25+ tools, MCP support, and workflow automation.">
   <meta name="twitter:image" content="https://zubo.bot/og-image.png">
   <meta name="twitter:creator" content="@thomaskanze">
   <link rel="preconnect" href="https://fonts.googleapis.com">
@@ -87,7 +87,7 @@
       <h1>Getting Started with Zubo</h1>
 
       <p>
-        Zubo is a self-hosted AI agent with persistent semantic memory, multi-channel presence across Telegram, Discord, Slack, WhatsApp, Signal, and Web Chat, over 35 smart built-in tools, sub-agent delegation, and workflow automation. It runs entirely on your machine &mdash; your data never leaves your infrastructure. Whether you need a personal assistant that remembers everything, an automation hub that connects your services, or an AI copilot embedded in every chat platform you use, Zubo handles it from a single process with a single configuration file.
+        Zubo is a self-hosted AI agent with persistent semantic memory, multi-channel presence across Telegram, Discord, Slack, WhatsApp, Signal, Email, and Web Chat, 25+ smart built-in tools, MCP compatibility, sub-agent delegation, and workflow automation. It runs entirely on your machine &mdash; your data never leaves your infrastructure. Whether you need a personal assistant that remembers everything, an automation hub that connects your services, or an AI copilot embedded in every chat platform you use, Zubo handles it from a single process with a single configuration file.
       </p>
 
       <!-- ================================================================ -->
@@ -110,6 +110,7 @@
         <li><strong>Slack</strong> &mdash; app-level integration using Bot Token and App Token, with user allow-listing by Slack member ID.</li>
         <li><strong>WhatsApp</strong> &mdash; via the WhatsApp Web multi-device API, with phone number allow-listing.</li>
         <li><strong>Signal</strong> &mdash; via signal-cli, with phone number allow-listing.</li>
+        <li><strong>Email</strong> &mdash; bidirectional email channel via IMAP/SMTP. Zubo polls for new messages and replies automatically. Supports sender allow-listing and configurable polling intervals.</li>
         <li><strong>Web Chat</strong> &mdash; built-in web dashboard with real-time chat, served on a configurable port.</li>
         <li>All channels share the same memory, personality, tools, and conversation history. A fact learned in Telegram is immediately available in Discord.</li>
       </ul>
@@ -121,11 +122,15 @@
         <li><strong>Hybrid search</strong> that combines 60% vector similarity (cosine) with 40% FTS5 BM25 scoring for best-of-both-worlds recall.</li>
         <li><strong>Automatic chunking with overlap</strong> for long documents, ensuring context is preserved across chunk boundaries.</li>
         <li><strong>Document ingestion</strong> supporting PDF, DOCX, TXT, CSV, JSON, YAML, and common code file formats. Upload a document and Zubo indexes it into searchable memory automatically.</li>
+        <li><strong>Knowledge Graph</strong> &mdash; structured entity-relationship memory. The agent can store and query entities (people, projects, concepts, organizations) with typed relationships (works_at, knows, manages, part_of). Entity context is automatically injected into conversations when relevant names are mentioned.</li>
       </ul>
 
       <h3>Tools</h3>
       <ul>
-        <li><strong>35+ built-in tools</strong> including shell command execution, file I/O (read, write, list, search), web search (DuckDuckGo, Brave, Google), HTTP requests, memory read/write/search, secrets management, and cron job scheduling.</li>
+        <li><strong>25+ built-in tools</strong> including shell command execution, file I/O (read, write, list, search), web search (DuckDuckGo, Brave, Google), HTTP requests, memory read/write/search, secrets management, cron job scheduling, code interpreter, image generation, knowledge graph, and webhook management.</li>
+        <li><strong>Code Interpreter</strong> &mdash; execute Python, JavaScript, or TypeScript code in a sandboxed subprocess with timeout enforcement and output capture. Runs code via Bun (JS/TS) or python3 (Python) with process isolation.</li>
+        <li><strong>Image Generation</strong> &mdash; generate images using DALL-E 3, fal.ai (Flux), or Together AI. Images are saved locally and served through the web dashboard. Supports custom sizes and styles.</li>
+        <li><strong>Webhooks</strong> &mdash; create webhook endpoints that receive events from external services (GitHub, Stripe, CI/CD). Supports HMAC signature verification (X-Hub-Signature-256). Events trigger agent actions automatically.</li>
         <li>Tools are permission-gated with three levels: <code>safe</code>, <code>sensitive</code>, and <code>dangerous</code>. Dangerous operations require explicit confirmation tokens.</li>
       </ul>
 
@@ -160,6 +165,14 @@
         <li><strong>Twitter</strong> &mdash; post tweets, read timelines.</li>
       </ul>
 
+      <h3>MCP (Model Context Protocol)</h3>
+      <ul>
+        <li><strong>Connect to external MCP servers</strong> and expose their tools alongside Zubo's built-in tools. Any MCP-compatible tool server works out of the box.</li>
+        <li><strong>JSON-RPC over stdio</strong> &mdash; each MCP server runs as a subprocess. Zubo handles the full protocol: initialization, tool discovery, and tool execution.</li>
+        <li><strong>Automatic tool registration</strong> &mdash; on boot, Zubo connects to each configured MCP server, discovers available tools, and registers them with a server-name prefix to avoid conflicts (e.g., <code>filesystem__read_file</code>).</li>
+        <li><strong>Configure in JSON</strong> &mdash; add MCP servers via the <code>mcp.servers</code> config array. Example: <code>{ "name": "filesystem", "command": "npx", "args": ["-y", "@modelcontextprotocol/server-filesystem", "/tmp"] }</code></li>
+      </ul>
+
       <h3>Scheduling</h3>
       <ul>
         <li><strong>Natural language scheduling</strong> &mdash; describe schedules in plain English like &ldquo;every weekday at 9am&rdquo; or &ldquo;twice a day&rdquo;. Zubo parses natural language into cron expressions automatically.</li>
@@ -200,21 +213,69 @@
       <!-- ================================================================ -->
       <h2 id="installation">Installation</h2>
 
+      <p>Choose the installation method that fits your platform.</p>
+
+      <h3 id="install-mac-linux">macOS &amp; Linux</h3>
+
       <p>Zubo requires <strong>Bun v1.0+</strong> as its runtime. If you do not have Bun installed, visit <a href="https://bun.sh" target="_blank" rel="noopener">bun.sh</a> and run the one-line installer.</p>
 
-      <h3>Step 1: Install Zubo</h3>
+      <p><strong>One-liner (recommended):</strong></p>
+
+<pre><code>curl -fsSL https://zubo.bot/install.sh | bash</code></pre>
 
-      <p>Install Zubo globally using Bun (recommended) or npm:</p>
+      <p>This installs Bun (if needed), installs Zubo globally, and runs the setup wizard automatically.</p>
+
+      <p><strong>Manual install:</strong></p>
 
 <pre><code># Using Bun (recommended)
 bun add -g zubo
 
 # Or using npm
-npm i -g zubo</code></pre>
+npm i -g zubo
+
+# Then run the setup wizard
+zubo setup</code></pre>
+
+      <h3 id="install-windows">Windows (via WSL)</h3>
+
+      <p>Zubo runs on Windows through the Windows Subsystem for Linux (WSL). This gives you full compatibility with zero workarounds.</p>
+
+      <p><strong>Step 1:</strong> Install WSL (requires Windows 10 version 2004+ or Windows 11):</p>
+
+<pre><code>wsl --install</code></pre>
+
+      <p>Restart your computer when prompted. This installs Ubuntu by default.</p>
+
+      <p><strong>Step 2:</strong> Open the Ubuntu terminal from the Start Menu and install Zubo:</p>
+
+<pre><code>curl -fsSL https://zubo.bot/install.sh | bash</code></pre>
+
+      <p>That's it. Zubo runs inside WSL with full access to all features. The web dashboard is accessible from your Windows browser at <code>http://localhost:3000</code>.</p>
+
+      <h3 id="install-docker">Docker</h3>
+
+      <p>Run Zubo as a container with no local dependencies.</p>
+
+<pre><code># Clone the repo
+git clone https://github.com/apwn/zubo.git &amp;&amp; cd zubo
+
+# Copy the example env and add your API key
+cp .env.example .env
+
+# Start Zubo
+docker compose up -d</code></pre>
+
+      <p>The container exposes port <code>8787</code> for the web dashboard and mounts <code>~/.zubo</code> for persistent data. Stop with <code>docker compose down</code>.</p>
+
+      <p>To configure, edit <code>.env</code> or mount your own <code>config.json</code>:</p>
+
+<pre><code>volumes:
+  - ~/.zubo:/root/.zubo
+  - ./my-config.json:/root/.zubo/config.json</code></pre>
 
-      <p>This places the <code>zubo</code> binary on your PATH, making it available from any terminal session.</p>
+      <h3 id="setup-wizard">The Setup Wizard</h3>
 
-      <h3>Step 2: Run the Setup Wizard</h3>
+      <p>After installing (on any platform), run the setup wizard:</p>
 
 <pre><code>zubo setup</code></pre>
 
@@ -228,7 +289,7 @@ npm i -g zubo</code></pre>
         <li>Create a default <code>SYSTEM.md</code> system prompt that you can customize later.</li>
       </ul>
 
-      <h3>Step 3: Start Zubo</h3>
+      <h3 id="starting">Starting Zubo</h3>
 
 <pre><code># Start in foreground (logs to stdout)
 zubo start