Package Exports

leedab
leedab/dist/index.js

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (leedab) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

LeedAB

Your enterprise AI agent. Local-first, private by default.

LeedAB deploys an AI agent on your own hardware — Mac, Linux, or VPS. It connects to your team's messaging apps, remembers context across conversations, handles tasks with browser automation, and keeps every byte of data on your machine.

Quick Start

mkdir my-agent && cd my-agent
npx leedab onboard
npx leedab start

That's it. The onboarding wizard walks you through:

LLM provider — Anthropic, OpenAI, Gemini, DeepSeek, Bedrock, or OpenRouter
Channels — Telegram, WhatsApp, Microsoft Teams
Credential vault — encrypted storage for service logins the agent uses

What You Get

Multi-channel messaging — your team talks to the agent on Telegram, WhatsApp, or Teams. Each user gets an isolated session with their own context.

Persistent memory — the agent remembers across conversations. Daily notes, long-term memory, and workspace files survive restarts.

Credential vault — AES-256-GCM encrypted storage. The agent logs into services (Gmail, CRM, carrier portals) using stored credentials via browser automation.

Dashboard — web console at localhost:3000 showing agent status, channel health, session history, and team management.

Your data stays local — files, credentials, memory, and logs live on your machine. Messages route through your channel provider and prompts go to your AI provider, but nothing is stored externally.

Commands

leedab onboard          Interactive setup wizard
leedab start            Start the agent and dashboard
leedab terminal         Chat with the agent in terminal
leedab stop             Stop the agent
leedab configure model  Change LLM provider or model
leedab vault add <svc>  Store credentials for a service
leedab pairing add      Add users to channel allowlist
leedab team add <name>  Add a team member
leedab sessions         View conversation sessions
leedab --help           All commands

Requirements

Node.js 22+
An LLM API key (Anthropic, OpenAI, etc.)
At least one messaging channel (Telegram is the easiest to set up)

How It Works

LeedAB runs an AI agent gateway on your machine with enterprise onboarding, a management dashboard, encrypted credential vault, team roles, and audit logging.

┌─────────────────────────────────────────────┐
│  LeedAB                                     │
│  ┌─────────┐ ┌──────────┐ ┌──────────────┐  │
│  │ Onboard │ │Dashboard │ │ Vault/Audit  │  │
│  └─────────┘ └──────────┘ └──────────────┘  │
│  ┌─────────────────────────────────────────┐ │
│  │  Agent Gateway                          │ │
│  │  Sessions │ Skills │ Memory │ Browser   │ │
│  └─────────────────────────────────────────┘ │
│  ┌──────────┐ ┌──────────┐ ┌─────────────┐  │
│  │ Telegram │ │ WhatsApp │ │   Teams     │  │
│  └──────────┘ └──────────┘ └─────────────┘  │
└─────────────────────────────────────────────┘

Security

All data stored locally in .leedab/ (gitignored)
Credentials encrypted with AES-256-GCM
Per-channel allowlists control who can message the agent
DM policies: open, pairing (approval required), or closed
Team roles: admin, operator, viewer
Full audit log of every interaction

License

Proprietary. See LICENSE for details.