OwnPilot

Privacy-first personal AI assistant platform with autonomous agents, tool orchestration, multi-provider support, and multi-channel communication.

Self-hosted. Your data stays yours.

Features

AI & Agents

100+ AI Provider Configs - OpenAI, Anthropic, Google, DeepSeek, Groq, xAI, Mistral, Together AI, Fireworks, Perplexity, OpenRouter, Cohere, Azure, NVIDIA, Hugging Face, and many more
Local AI Support - Ollama, LM Studio, and custom OpenAI-compatible endpoints
29 Pre-configured Agents - Code Assistant, Writing Assistant, Research, Data Analyst, and more across 8 categories
Smart Provider Routing - Cheapest, fastest, smartest, balanced, or fallback strategies
Streaming Responses - Server-Sent Events (SSE) for real-time streaming

Tools

148+ Built-in Tools across 20 categories (personal data, files, code, web, email, media, AI/NLP, finance, automation, utilities)
Meta-tool Proxy - Only 3 tools sent to the LLM (search_tools, get_tool_help, use_tool); all 148+ tools remain available via dynamic discovery
Batch Operations - batch_use_tool for multiple tool calls in one request, search_tools with include_params for faster workflows
Custom Tools - Create new tools at runtime via LLM (sandboxed JavaScript)
Tool Limits - Automatic parameter capping to prevent unbounded queries
Search Tags - 150+ English keywords for natural language tool discovery

Personal Data

Notes, Tasks, Bookmarks, Contacts, Calendar, Expenses - Full CRUD with categories, tags, and search
Productivity - Pomodoro timer with sessions/settings/stats, habit tracker with streaks, quick capture inbox
Memories - Long-term persistent memory (facts, preferences, events, skills) with importance scoring
Goals - Goal creation, decomposition, progress tracking, next-action recommendations
Custom Data Tables - Create your own structured data types

Autonomy & Automation

5 Autonomy Levels - Manual, Assisted, Supervised, Autonomous, Full
Triggers - Schedule-based (cron), event-driven, condition-based, webhook
Plans - Multi-step autonomous execution with checkpoints and rollback
Risk Assessment - Automatic risk scoring for tool executions

Communication

Web UI - React 19 + Vite + Tailwind CSS 4 with dark mode
Telegram Bot - Full bot integration with user/chat filtering
Discord Bot - Multi-guild support with DMs
Slack Bot - Socket Mode with thread support
REST API - 35 route modules, full CRUD for all entities

Security

Encrypted Credential Storage - AES-256-GCM + PBKDF2 (600K iterations)
PII Detection & Redaction - Automatic pattern-based detection
Sandboxed Code Execution - Isolated VM with resource limits
Authentication - None, API Key, or JWT modes
Rate Limiting - Configurable with burst support and soft mode

Architecture Overview

                         ┌──────────────┐
                         │   Web UI     │  React 19 + Vite
                         │  (Port 3000) │  + Tailwind CSS 4
                         └──────┬───────┘
                                │
              ┌─────────────────┼─────────────────┐
              │                 │                  │
     ┌────────┴────────┐       │        ┌─────────┴──────────┐
     │  Telegram Bot   │       │        │   Discord / Slack  │
     │   (Channels)    │       │        │     (Channels)     │
     └────────┬────────┘       │        └─────────┬──────────┘
              │                │                  │
              └────────┬───────┘──────────────────┘
                       │
              ┌────────▼────────┐
              │    Gateway      │  Hono API Server
              │  (Port 8080)    │  35 Route Modules
              ├─────────────────┤
              │  Service Layer  │  Business Logic
              │  Agent Engine   │  Tool Orchestration
              │  Provider Router│  Smart Model Selection
              │  Autonomy       │  Risk Assessment
              │  EventBus       │  Typed Event System
              ├─────────────────┤
              │     Core        │  Tools, Providers, Types
              │  148+ Tools     │  100+ Provider Configs
              │  Sandbox, Crypto│  Privacy, Audit
              └────────┬────────┘
                       │
              ┌────────▼────────┐
              │   PostgreSQL    │  47 Tables
              │  (Port 25432)   │  Conversations, Personal Data,
              │                 │  Memories, Goals, Triggers, Plans
              └─────────────────┘

Quick Start

Prerequisites

Node.js >= 22.0.0
pnpm >= 9.0.0
Docker (for PostgreSQL)

Setup

# Clone and install
git clone https://github.com/ownpilot/ownpilot.git
cd ownpilot
pnpm install

# Start PostgreSQL
docker compose -f docker-compose.db.yml up -d

# Configure
cp .env.example .env
# Edit .env - add at least one API key (OPENAI_API_KEY, ANTHROPIC_API_KEY, etc.)

# Seed database with default agents
pnpm --filter @ownpilot/gateway seed

# Start development (gateway + ui)
pnpm dev

# UI: http://localhost:3000
# API: http://localhost:8080

Secure Credential Storage (Recommended)

Instead of storing API keys in .env, use the encrypted credential store:

# Create encrypted store
ownpilot setup

# Add API keys (encrypted with AES-256-GCM)
ownpilot config set openai-api-key
ownpilot config set anthropic-api-key

# Start with auto-unlock
OWNPILOT_PASSWORD=your-master-password ownpilot start

Credentials are stored in ~/.ownpilot/credentials.enc with AES-256-GCM encryption and PBKDF2 key derivation (600K iterations).

Project Structure

ownpilot/
├── packages/
│   ├── core/                    # AI runtime, providers, tools, types
│   │   ├── src/
│   │   │   ├── agent/           # Agent engine, orchestrator, types
│   │   │   │   ├── providers/   # 100+ provider configs (JSON), synced from models.dev
│   │   │   │   ├── tools/       # 23 tool modules (148+ tools)
│   │   │   │   └── types.ts     # Agent, provider, ToolProvider, ToolMiddleware types
│   │   │   ├── events/          # EventBus - typed event system
│   │   │   ├── memory/          # Conversation & personal memory
│   │   │   ├── sandbox/         # Isolated code execution (VM)
│   │   │   ├── privacy/         # PII detection & redaction
│   │   │   ├── crypto/          # AES-256-GCM encryption, keychain
│   │   │   ├── audit/           # Audit logging & compliance
│   │   │   ├── costs/           # API cost tracking
│   │   │   ├── plugins/         # Plugin system (worker threads)
│   │   │   ├── services/        # Config center, service registry
│   │   │   └── types/           # Branded types, Result<T,E>
│   │   └── package.json
│   │
│   ├── gateway/                 # Hono API server
│   │   ├── src/
│   │   │   ├── routes/          # 35 route modules
│   │   │   ├── services/        # 16 business logic services, tool providers
│   │   │   ├── db/
│   │   │   │   ├── repositories/  # 30 data access repositories
│   │   │   │   ├── migrations/    # PostgreSQL schema (47 tables)
│   │   │   │   └── seeds/         # Default agents, config services
│   │   │   ├── channels/        # Discord, Slack adapters
│   │   │   ├── triggers/        # Trigger engine
│   │   │   ├── plans/           # Plan executor
│   │   │   ├── ws/              # WebSocket handler
│   │   │   └── middleware/      # Auth, rate limiting, CORS
│   │   ├── data/seeds/          # default-agents.json
│   │   └── package.json
│   │
│   ├── ui/                      # React 19 web interface
│   │   ├── src/
│   │   │   ├── pages/           # 28+ page components
│   │   │   ├── components/      # 22+ reusable components
│   │   │   ├── hooks/           # Custom React hooks
│   │   │   ├── types/           # UI type definitions
│   │   │   ├── App.tsx          # Route definitions
│   │   │   └── main.tsx         # Entry point
│   │   └── package.json
│   │
│   ├── channels/                # Messaging channel adapters
│   │   ├── src/
│   │   │   ├── telegram/        # Telegram Bot API
│   │   │   ├── manager.ts       # Channel orchestration
│   │   │   └── types/           # Channel type definitions
│   │   └── package.json
│   │
│   └── cli/                     # Command-line tools
│       ├── src/
│       │   ├── commands/        # server, bot, start, config, workspace, channel
│       │   └── index.ts         # CLI entry (Commander.js)
│       └── package.json
│
├── docker-compose.db.yml        # PostgreSQL only
├── docker-compose.yml           # Full stack (gateway + ui + db)
├── Dockerfile                   # Multi-stage production build
├── turbo.json                   # Turborepo task config
├── tsconfig.base.json           # Shared TypeScript config
└── package.json                 # Monorepo root

Packages

Core (`@ownpilot/core`)

The core runtime library. Contains all AI provider integrations, tool definitions, agent types, and security primitives.

Key Modules:

Module	Description
`agent/providers/`	100+ provider config JSONs synced from models.dev
`agent/tools/`	23 tool modules with 148+ tool definitions and executors
`agent/types.ts`	Agent config, provider types, ToolProvider, ToolMiddleware
`events/`	EventBus - typed event system with wildcard subscriptions
`memory/`	Conversation memory and personal memory stores
`sandbox/`	Isolated VM code execution with resource limits
`privacy/`	PII detection patterns (email, phone, SSN, credit card, etc.)
`crypto/`	AES-256-GCM encryption, PBKDF2 key derivation, vault
`audit/`	Event logging, verification, compliance
`costs/`	Per-provider/model cost calculation
`plugins/`	Worker-thread isolated plugin runtime
`services/`	Config center, API service registry

Gateway (`@ownpilot/gateway`)

The API server built on Hono. Handles all HTTP/WebSocket communication, database operations, agent execution, and channel management.

Route Modules (35):

Category	Routes
Chat & Agents	`chat.ts`, `agents.ts`
AI Configuration	`models.ts`, `providers.ts`, `model-configs.ts`, `local-providers.ts`
Personal Data	`personal-data.ts`, `personal-data-tools.ts`, `memories.ts`, `goals.ts`, `expenses.ts`, `custom-data.ts`
Productivity	`productivity.ts` (Pomodoro, Habits, Captures)
Automation	`triggers.ts`, `plans.ts`, `autonomy.ts`
Tools & Plugins	`tools.ts`, `custom-tools.ts`, `plugins.ts`
Channels	`channels.ts`
Configuration	`settings.ts`, `config-services.ts`, `media-settings.ts`
Integration	`integrations.ts`, `auth.ts`
System	`health.ts`, `dashboard.ts`, `costs.ts`, `audit.ts`, `debug.ts`, `database.ts`, `profile.ts`, `workspaces.ts`, `file-workspaces.ts`

Services (16): GoalService, MemoryService, CustomDataService, TriggerService, PlanService, DashboardService, ToolExecutor, ToolOverrides, ToolProviders, GmailExecutors, MediaExecutors, ConfigCenter, ConfigTools, ApiServiceRegistrar, LocalDiscovery, ToolSource.

Database Repositories (30):

agents, conversations, messages, chat, tasks, notes, bookmarks, calendar, contacts, memories, goals, triggers, plans, expenses, custom-data, custom-tools, plugins, channels, channel-messages, costs, logs, settings, workspaces, model-configs, media-settings, oauth-integrations, local-providers, config-services, pomodoro, habits, captures.

UI (`@ownpilot/ui`)

Modern web interface built with React 19, Vite 6, and Tailwind CSS 4.

Tech Stack:

React 19.0.0 with hooks and Context API
React Router DOM 7.1.3
Vite 6.0.11
Tailwind CSS 4.0.6
Prism React Renderer (syntax highlighting)
Lucide React (icons)

Pages (28+):

Page	Description
Chat	Main AI conversation interface with streaming, progress events, tool execution display
Dashboard	Overview with stats cards, AI briefing, timeline, quick actions
Tasks	Task management with priorities, due dates, categories
Notes	Note-taking with markdown support
Calendar	Event scheduling and viewing
Contacts	Contact directory with relationships
Bookmarks	URL bookmarking with tags and categories
Expenses	Financial tracking with categories
Memories	Browse and manage AI long-term memories
Goals	Goal tracking with progress and decomposition
Data Browser	Universal data browser for all personal data types
Custom Data	User-defined data tables and records
Agents	Agent selection and configuration
Tools	Tool browser with categories
Custom Tools	Create/manage LLM-created tools
Models	AI model browser across providers
Costs	API usage and spending dashboard
Triggers	Automation trigger management
Plans	Multi-step plan management
Autonomy	Autonomy level controls, budget limits, tool permissions
Plugins	Plugin management
Workspaces	Multi-workspace support
Config Center	Dynamic service configuration
API Keys	Provider API key management
Logs	Request logs and debug logs
System	System info, database backup/restore, sandbox status
Profile	User profile settings
Settings	Settings hub (Providers, AI Models, Integrations, Media)

Key Components:

Layout - 3-column layout (sidebar, content, stats panel)
MessageList - Chat messages with code blocks, tool calls, traces
ChatInput - Message input with send/stop controls
ToolExecutionDisplay - Real-time tool execution results
TraceDisplay - Detailed execution trace viewer
CodeBlock - Syntax-highlighted code with copy button
ConfirmDialog - Styled modal dialogs (replaces native confirm/alert)
DynamicConfigForm - Schema-driven configuration forms
AIBriefingCard - AI-generated daily briefing
TimelineView - Timeline visualization

Custom Hooks:

useChatStore - Global chat state (provider, model, messages, streaming)
useTheme - Dark/light/system theme with localStorage
useWebSocket - WebSocket connection management
useDialog - Async confirm/alert dialogs (Promise<boolean>)

Channels (`@ownpilot/channels`)

Communication channel adapters for messaging platforms.

Channel	Features
Telegram	Bot API, message/photo/document/audio handling, user/chat filtering, parse modes (HTML, Markdown)
Discord	Discord.js, multi-guild, DM support, message content intents, partial message reconstruction
Slack	@slack/bolt, Socket Mode, thread support, file attachments, emoji reactions

All channels implement a common ChannelHandler interface with start(), stop(), sendMessage(), and onMessage() methods.

CLI (`@ownpilot/cli`)

Command-line interface built with Commander.js.

ownpilot setup                    # Create encrypted credential store
ownpilot start                    # Start all services (server + bot)
ownpilot server                   # Start HTTP API server only
ownpilot bot                      # Start Telegram bot only

# Credential management
ownpilot config set <key>         # Add encrypted credential
ownpilot config get <key>         # Retrieve credential
ownpilot config delete <key>      # Delete credential
ownpilot config list              # List all credentials
ownpilot config change-password   # Change master password

# Workspace management
ownpilot workspace list           # List workspaces
ownpilot workspace create <name>  # Create workspace
ownpilot workspace delete <name>  # Delete workspace
ownpilot workspace switch <name>  # Switch active workspace

# Channel management
ownpilot channel list             # List channels
ownpilot channel add <type>       # Add channel (telegram/discord/slack)
ownpilot channel connect <id>     # Connect channel
ownpilot channel disconnect <id>  # Disconnect channel

AI Providers

OwnPilot supports 100+ AI provider configurations via OpenAI-compatible API format, Anthropic native API, and Google Gemini API.

Supported Providers

Provider	Type	Notable Models	Env Variable
OpenAI	Native	GPT-5, GPT-5.1, o3, o4-mini, GPT-4o	`OPENAI_API_KEY`
Anthropic	Native	Claude Opus 4.5, Claude Sonnet 4.5, Claude Haiku 4.5	`ANTHROPIC_API_KEY`
Google	Native	Gemini 2.0 Pro, Gemini 2.0 Flash	`GOOGLE_API_KEY`
DeepSeek	OpenAI-compat	DeepSeek V3.2, DeepSeek Reasoner	`DEEPSEEK_API_KEY`
Groq	OpenAI-compat	Llama 4 Maverick, Mixtral 8x7B	`GROQ_API_KEY`
xAI	OpenAI-compat	Grok 3, Grok 3 Mini	`XAI_API_KEY`
Mistral	OpenAI-compat	Mistral Large 3, Devstral 2, Codestral	`MISTRAL_API_KEY`
Together AI	OpenAI-compat	Llama 4, Qwen3-72B	`TOGETHER_API_KEY`
Fireworks AI	OpenAI-compat	Llama 4, Mistral Large 3	`FIREWORKS_API_KEY`
Perplexity	OpenAI-compat	Sonar Pro, Sonar Reasoning	`PERPLEXITY_API_KEY`
OpenRouter	OpenAI-compat	200+ models via single API	`OPENROUTER_API_KEY`
Ollama	Local	Llama 4, Qwen3, Mistral, CodeLlama	None (local)
LM Studio	Local	Any GGUF model	None (local)

Plus 85+ more providers including Azure, AWS Bedrock, Google Vertex, NVIDIA, Hugging Face, Cerebras, Scaleway, OVHcloud, and more.

Provider Routing Strategies

The provider router can automatically select the best model based on:

Strategy	Description
`cheapest`	Minimize API costs
`fastest`	Minimize latency
`smartest`	Best quality/reasoning
`balanced`	Cost + quality balance (default)
`fallback`	Try providers sequentially until one succeeds

Agent System

Pre-configured Agents (29)

Agents are AI assistants with specific system prompts, tool assignments, and model preferences.

Category	Agents
Core (2)	Orchestrator, General Assistant
Technical (6)	Code Assistant, DevOps Engineer, Database Expert, Security Analyst, Mobile Developer, API Designer
Content (5)	Writing Assistant, Creative Writer, Technical Writer, Summarizer, Translator
Professional (5)	Task Manager, Career Coach, Personal Coach, Legal Assistant, Finance Advisor
Analysis (3)	Data Analyst, Research Assistant, Business Analyst
Creative (3)	UX Designer, Image Prompt Creator, Video Scriptwriter
Education (3)	Math Tutor, Language Tutor, Study Coach
Specialized (2)	Email Composer, Meeting Assistant

Agent Configuration

Each agent has:

{
  name: string               // Display name
  systemPrompt: string       // Custom instructions
  provider: string           // AI provider (or 'default')
  model: string              // Model ID (or 'default')
  config: {
    maxTokens: number        // Max response tokens
    temperature: number      // Creativity (0-2)
    maxTurns: number         // Max conversation turns
    maxToolCalls: number     // Max tool calls per turn
    tools?: string[]         // Specific tool names
    toolGroups?: string[]    // Tool group names
  }
}

Agent Capabilities

Tool Orchestration - Automatic tool calling with multi-step planning
Memory Injection - Relevant memories automatically included in context
Dynamic Prompts - Context-aware system prompt enhancement
Error Handling - Automatic retries with graceful degradation

Tool System

Overview

OwnPilot has 148+ tools organized into 20 categories. Rather than sending all tool definitions to the LLM (which would consume too many tokens), OwnPilot uses a meta-tool proxy pattern:

search_tools - Find tools by keyword (supports include_params for inline parameter schemas)
get_tool_help - Get detailed help for a specific tool (supports batch lookup)
use_tool - Execute a tool with automatic parameter validation and limit enforcement
batch_use_tool - Execute multiple tools in a single call for faster workflows

Tool Categories

Category	Tools	Examples
Tasks	6	add_task, list_tasks, complete_task, update_task, delete_task, batch_add_tasks
Notes	5	add_note, list_notes, update_note, delete_note, batch_add_notes
Calendar	4	add_calendar_event, list_calendar_events, delete_calendar_event, batch_add
Contacts	5	add_contact, list_contacts, update_contact, delete_contact, batch_add
Bookmarks	4	add_bookmark, list_bookmarks, delete_bookmark, batch_add_bookmarks
Custom Data	11	create_custom_table, add_custom_record, search_custom_records, ...
File System	8	read_file, write_file, list_directory, search_files, copy_file, ...
PDF	3	read_pdf, create_pdf, pdf_info
Code Execution	5	execute_javascript, execute_python, execute_shell, compile_code, package_manager
Git	7	git_status, git_diff, git_log, git_commit, git_add, git_branch, git_checkout
Web & API	4	http_request, fetch_web_page, search_web, json_api
Email	6	send_email, list_emails, read_email, search_emails, reply_email, delete_email
Image	5	analyze_image, generate_image, edit_image, image_variation, resize_image
Audio	5	text_to_speech, speech_to_text, translate_audio, audio_info, split_audio
Translation	4	translate_text, detect_language, list_languages, batch_translate
Data Extraction	4	extract_structured_data, extract_entities, extract_table_data, summarize_text
Vector Search	7	create_embedding, semantic_search, upsert_vectors, similarity_score, ...
Finance	7	add_expense, parse_receipt, query_expenses, expense_summary, export_expenses, ...
Scheduler	6	create_scheduled_task, list_scheduled_tasks, trigger_task, ...
Weather	2	get_weather, get_weather_forecast
Memory	7	remember, recall, forget, list_memories, boost_memory, memory_stats, batch_remember
Goals	8	create_goal, list_goals, decompose_goal, get_next_actions, complete_step, ...
Dynamic Tools	4	create_tool, list_custom_tools, delete_custom_tool, toggle_custom_tool
Utilities	21	calculate, statistics, convert_units, generate_uuid, hash_text, regex, parse_csv, ...

Tool Limits

Automatic parameter capping prevents unbounded queries from the LLM:

Tool	Parameter	Max	Default
`list_emails`	limit	50	20
`search_emails`	limit	100	50
`list_tasks`	limit	50	20
`list_notes`	limit	50	20
`list_calendar_events`	limit	50	20
`list_contacts`	limit	50	20
`list_goals`	limit	30	10
`get_next_actions`	limit	20	5
`query_expenses`	limit	100	50
`semantic_search`	topK	50	10
`search_web`	maxResults	20	10
`search_files`	maxResults	100	50

Custom Tools (LLM-Created)

The AI can create new tools at runtime:

LLM calls create_tool with name, description, parameters, and JavaScript code
Tool is sandboxed and stored in the database
Tool is available to all agents via use_tool
Tools can be enabled/disabled, require approval, and have permission controls

Personal Data

Entity Types

Entity	Features
Tasks	Title, description, status (pending/in_progress/completed/cancelled), priority (low/normal/high/urgent), due date, category, tags, subtasks, recurrence
Notes	Title, content (markdown), category, tags, pinned, archived, color
Bookmarks	URL, title, description, category, tags, favorite, visit count
Calendar Events	Title, description, location, start/end time, all-day, timezone, recurrence, attendees, reminders
Contacts	Name, email, phone, company, job title, birthday, address, relationship, tags, social links, custom fields
Expenses	Amount, category, description, date, tags, receipt parsing
Custom Data	User-defined tables with dynamic schemas

Memory System

Persistent long-term memory for the AI assistant:

Memory Type	Description
`fact`	Factual information about the user
`preference`	User preferences and settings
`conversation`	Key conversation takeaways
`event`	Important events and milestones
`skill`	Learned capabilities

Memories have importance scoring (0-1), are automatically injected into agent context, and support semantic recall with relevance matching.

Goals System

Hierarchical goal tracking with decomposition:

Create goals with title, description, priority (1-10), due date
Decompose goals into sub-goals and actionable steps
Track progress (0-100%) with status (active/paused/completed/abandoned)
Get next actions - AI recommends what to do next
Complete steps and auto-update parent goal progress

Autonomy & Automation

Autonomy Levels

Level	Name	Description
0	Manual	Always ask before any action
1	Assisted	Suggest actions, wait for approval (default)
2	Supervised	Auto-execute low-risk, ask for high-risk
3	Autonomous	Execute all actions, notify user
4	Full	Fully autonomous, minimal notifications

Triggers

Proactive automation with 4 trigger types:

Type	Description	Example
`schedule`	Cron-based timing	"Every Monday at 9am, summarize my week"
`event`	Fired on data changes	"When a new task is added, notify me"
`condition`	IF-THEN rules	"If expenses > $500/day, alert me"
`webhook`	External HTTP triggers	"When GitHub webhook fires, create a task"

Plans

Multi-step autonomous execution:

Step types: tool_call, llm_decision, user_input, condition, parallel, loop, sub_plan
Status tracking: pending, running, paused, completed, failed, cancelled
Checkpoints: Save state for rollback on failure
Linked to goals and triggers

Database

Supported Databases

Database	Status	Notes
PostgreSQL 16+	Primary	Recommended for production
SQLite	Legacy	Development/single-user

Schema (47 Tables)

Core Tables:

conversations, messages - Chat history
agents - Agent configurations
settings - Key-value settings store
request_logs - API request/response logging
costs - Token usage and cost tracking

Personal Data Tables:

tasks, notes, bookmarks, calendar_events, contacts
projects, reminders, captures (quick capture inbox)
expenses

Productivity Tables:

pomodoro_sessions, pomodoro_settings - Pomodoro timer
habits, habit_logs - Habit tracking

Autonomous AI Tables:

memories - Long-term memory with embeddings
goals, goal_steps - Goal hierarchy
triggers, trigger_history - Automation triggers
plans, plan_steps, plan_history - Multi-step plans

System Tables:

channels, channel_messages - Messaging channels
custom_tools - LLM-created tools
custom_data_tables, custom_data_records - User-defined data
plugins - Plugin state persistence
oauth_integrations - OAuth token storage
media_provider_settings - Media provider configuration
user_model_configs - Per-user model enable/disable
local_providers, local_models - Local AI provider management

Migration

# Run PostgreSQL migration
pnpm --filter @ownpilot/gateway migrate:postgres

# Dry run (preview only)
pnpm --filter @ownpilot/gateway migrate:postgres:dry

Security & Privacy

Credential Management

Two approaches for storing API keys:

Encrypted Vault (CLI) — ownpilot config set <key> stores credentials in ~/.ownpilot/credentials.enc with AES-256-GCM encryption and PBKDF2 key derivation (600K iterations). Recommended for CLI usage.
Database Settings (Gateway API) — POST /api/v1/settings/api-keys stores keys in PostgreSQL via the config services system. Keys are loaded into process.env at startup for provider SDK compatibility. Used by the Web UI.

PII Detection

Automatic detection of emails, phone numbers, SSNs, credit cards, and more
Configurable redaction before logging/storage
Enable via ENABLE_PII_REDACTION=true

Sandboxed Code Execution

Node.js VM module + Worker threads for isolation
Configurable resource limits (memory, CPU, timeout)
Permission-based access control (filesystem, network, process)
All executions are audit-logged

Authentication

None - No authentication (development only — not for production)
API Key - Static API keys in API_KEYS env var
JWT - JSON Web Token with configurable secret
Database Admin - Mutating database operations (/api/v1/database/*) require ADMIN_API_KEY env var and X-Admin-Key header, regardless of global auth config

Rate Limiting

Window-based rate limiting (default: 500 requests/minute)
Burst support (default: 750)
Soft mode: warn but don't block (recommended for personal use)
Fully configurable or disableable

API Reference

Core Endpoints

Method	Endpoint	Description
`POST`	`/api/v1/chat`	Send message to AI agent
`GET`	`/api/v1/chat/conversations`	List conversations
`GET`	`/api/v1/chat/conversations/:id`	Get conversation with messages

Agents

Method	Endpoint	Description
`GET`	`/api/v1/agents`	List all agents
`POST`	`/api/v1/agents`	Create new agent
`PUT`	`/api/v1/agents/:id`	Update agent
`DELETE`	`/api/v1/agents/:id`	Delete agent

AI Configuration

Method	Endpoint	Description
`GET`	`/api/v1/models`	List available models
`GET`	`/api/v1/providers`	List providers with status
`GET`	`/api/v1/tools`	List all tools

Personal Data

Method	Endpoint	Description
`GET/POST`	`/api/v1/tasks`	Tasks CRUD
`GET/POST`	`/api/v1/notes`	Notes CRUD
`GET/POST`	`/api/v1/bookmarks`	Bookmarks CRUD
`GET/POST`	`/api/v1/calendar`	Calendar events CRUD
`GET/POST`	`/api/v1/contacts`	Contacts CRUD
`GET/POST`	`/api/v1/expenses`	Expenses CRUD
`GET/POST`	`/api/v1/memories`	Memories CRUD
`GET/POST`	`/api/v1/goals`	Goals CRUD

Automation

Method	Endpoint	Description
`GET/POST`	`/api/v1/triggers`	Trigger management
`GET/POST`	`/api/v1/plans`	Plan management
`GET/PUT`	`/api/v1/autonomy`	Autonomy settings

System

Method	Endpoint	Description
`GET`	`/health`	Health check
`GET`	`/api/v1/dashboard`	Dashboard data
`GET`	`/api/v1/costs`	Cost tracking
`GET`	`/api/v1/chat/logs`	Request logs

Configuration

Environment Variables

# ─── Master Password ───────────────────────────────
OWNPILOT_PASSWORD=              # Auto-unlock encrypted credentials

# ─── AI Provider API Keys ─────────────────────────
OPENAI_API_KEY=sk-...
ANTHROPIC_API_KEY=sk-ant-...
GOOGLE_API_KEY=...
DEEPSEEK_API_KEY=...
GROQ_API_KEY=...
XAI_API_KEY=...
MISTRAL_API_KEY=...

# ─── Server ────────────────────────────────────────
PORT=8080
HOST=0.0.0.0
NODE_ENV=development
CORS_ORIGINS=http://localhost:3000,http://localhost:5173

# ─── Authentication ────────────────────────────────
AUTH_TYPE=none                   # none | api-key | jwt
API_KEYS=key1,key2              # For api-key auth
JWT_SECRET=your-secret          # For jwt auth (min 32 chars)
ADMIN_API_KEY=your-admin-key    # Required for database admin operations

# ─── Rate Limiting ─────────────────────────────────
RATE_LIMIT_DISABLED=false
RATE_LIMIT_WINDOW_MS=60000      # 1 minute window
RATE_LIMIT_MAX=500              # Max requests per window
RATE_LIMIT_BURST=750            # Allow temporary spikes
RATE_LIMIT_SOFT=true            # Warn but don't block

# ─── Database ──────────────────────────────────────
DB_TYPE=postgres                # postgres | sqlite
POSTGRES_HOST=localhost
POSTGRES_PORT=25432
POSTGRES_USER=ownpilot
POSTGRES_PASSWORD=ownpilot_secret
POSTGRES_DB=ownpilot
POSTGRES_POOL_SIZE=10
# Or use connection URL:
# DATABASE_URL=postgresql://ownpilot:secret@localhost:25432/ownpilot

# ─── Telegram Bot ──────────────────────────────────
TELEGRAM_BOT_TOKEN=123456789:ABC...
TELEGRAM_ALLOWED_USERS=         # Comma-separated user IDs
TELEGRAM_ALLOWED_CHATS=         # Comma-separated chat IDs

# ─── Autonomy ─────────────────────────────────────
DEFAULT_AUTONOMY_LEVEL=1        # 0=Manual, 1=Assisted, 2=Supervised, 3=Autonomous, 4=Full
ENABLE_PROACTIVE_TRIGGERS=false
TRIGGER_CHECK_INTERVAL=60000    # ms

# ─── Privacy & Security ───────────────────────────
ENABLE_PII_REDACTION=true
ENCRYPTION_KEY=                 # 32 bytes hex

# ─── Logging ───────────────────────────────────────
LOG_LEVEL=info
LOG_FORMAT=json

# ─── AI Agent ──────────────────────────────────────
SYSTEM_PROMPT=You are a helpful AI assistant.
DATA_DIR=./data

Deployment

Docker (Recommended)

# Full stack: PostgreSQL + Gateway + UI
docker compose up -d

# PostgreSQL only
docker compose -f docker-compose.db.yml up -d

# Build and run production image
docker build -t ownpilot .
docker run -p 8080:8080 --env-file .env ownpilot

The Dockerfile uses multi-stage builds:

Builder - node:22-alpine, installs pnpm, builds all packages
Production - node:22-alpine, copies only dist/ and prod dependencies

Docker Compose Services

Service	Image	Port	Description
`postgres`	postgres:16-alpine	25432	PostgreSQL database
`gateway`	ownpilot (built)	8080	API server
`ui`	ownpilot-ui (built)	3000	Web interface

Manual Deployment

# Build
pnpm build

# Start production server
pnpm start

# Or start individually
pnpm --filter @ownpilot/gateway start
pnpm --filter @ownpilot/ui preview

Development

Scripts

# ─── Root (Turborepo) ──────────────────────────────
pnpm dev              # Watch mode for all packages
pnpm build            # Build all packages
pnpm test             # Run all tests
pnpm test:watch       # Watch test mode
pnpm test:coverage    # Coverage reports
pnpm lint             # ESLint all packages
pnpm lint:fix         # Auto-fix lint issues
pnpm typecheck        # TypeScript type checking
pnpm clean            # Clear all build artifacts
pnpm format           # Prettier formatting
pnpm format:check     # Check formatting

# ─── Package-specific ──────────────────────────────
pnpm --filter @ownpilot/gateway seed                # Seed default agents
pnpm --filter @ownpilot/gateway seed:triggers-plans # Seed triggers & plans
pnpm --filter @ownpilot/gateway migrate:postgres    # Run DB migration
pnpm --filter @ownpilot/ui dev                      # UI dev server
pnpm --filter @ownpilot/ui build                    # Build UI

Tech Stack

Layer	Technology
Monorepo	pnpm workspaces + Turborepo
Language	TypeScript 5.9 (strict, ES2023, NodeNext)
API Server	Hono
Web UI	React 19 + Vite 6 + Tailwind CSS 4
Database	PostgreSQL 16 / SQLite
Testing	Vitest
Linting	ESLint
Formatting	Prettier
Git Hooks	Husky
Build	Turborepo with incremental caching
Container	Docker multi-stage

Testing

65 test files across all packages
1,075+ tests in the gateway package alone
Integration tests for all route modules
Unit tests for services, middleware, and core modules
Framework: Vitest with vi.mock for module-level mocking

Architecture Patterns

Pattern	Usage
Result<T, E>	Error handling throughout core
Repository	Data access abstraction (`IRepository<T>`, `StandardQuery`, `PaginatedResult`)
Strategy	Provider routing (cheapest/fastest/smartest)
Registry	Tool registration and discovery
EventBus	Typed event system with wildcard subscriptions (tool, resource, agent, system events)
Service Layer	Business logic services (GoalService, MemoryService, PlanService, TriggerService, etc.)
Tool Provider	Modular tool registration via `ToolProvider` interface and `ToolMiddleware`
Middleware	Hono request pipeline (auth, rate limiting, timing, circuit breaker)
Context + Hook	React state management (Chat, Theme, Dialog)

License

MIT

Name		Name	Last commit message	Last commit date
Latest commit History 144 Commits
.serena		.serena
docs		docs
packages		packages
scripts		scripts
.dockerignore		.dockerignore
.env.example		.env.example
.gitignore		.gitignore
.prettierrc		.prettierrc
Dockerfile		Dockerfile
README.md		README.md
docker-compose.db.yml		docker-compose.db.yml
docker-compose.yml		docker-compose.yml
models-dev-full.json		models-dev-full.json
ownpilot_logo.jpeg		ownpilot_logo.jpeg
package.json		package.json
pnpm-lock.yaml		pnpm-lock.yaml
pnpm-workspace.yaml		pnpm-workspace.yaml
start.ps1		start.ps1
start.sh		start.sh
tsconfig.base.json		tsconfig.base.json
turbo.json		turbo.json

ownpilot/OwnPilot

Folders and files

Latest commit

History

Repository files navigation

OwnPilot

Table of Contents

Features

AI & Agents

Tools

Personal Data

Autonomy & Automation

Communication

Security

Architecture Overview

Quick Start

Prerequisites

Setup

Secure Credential Storage (Recommended)

Project Structure

Packages

Core (@ownpilot/core)

Gateway (@ownpilot/gateway)

UI (@ownpilot/ui)

Channels (@ownpilot/channels)

CLI (@ownpilot/cli)

AI Providers

Supported Providers

Provider Routing Strategies

Agent System

Pre-configured Agents (29)

Agent Configuration

Agent Capabilities

Tool System

Overview

Tool Categories

Tool Limits

Custom Tools (LLM-Created)

Personal Data

Entity Types

Memory System

Goals System

Autonomy & Automation

Autonomy Levels

Triggers

Plans

Database

Supported Databases

Schema (47 Tables)

Migration

Security & Privacy

Credential Management

PII Detection

Sandboxed Code Execution

Authentication

Rate Limiting

API Reference

Core Endpoints

Agents

AI Configuration

Personal Data

Automation

System

Configuration

Environment Variables

Deployment

Docker (Recommended)

Docker Compose Services

Manual Deployment

Development

Scripts

Tech Stack

Testing

Architecture Patterns

License

About

Resources

Uh oh!

Stars

Watchers

Core (`@ownpilot/core`)

Gateway (`@ownpilot/gateway`)

UI (`@ownpilot/ui`)

Channels (`@ownpilot/channels`)

CLI (`@ownpilot/cli`)

Packages