Summary

Lemma Platform offers an open-source, local-first workspace unifying human and AI agent collaboration, moving beyond ephemeral chat. It addresses the challenge of managing persistent, stateful work that spans multiple tools and participants. The platform enables teams and individual developers to integrate AI agents into structured workflows, providing a persistent environment for tasks, approvals, and data management, thereby enhancing productivity and operational clarity.

How It Works

The core concept is the "pod," a self-contained directory housing all workspace components: typed tables, files, agents, workflows, permissions, and applications. This file-based, local-first architecture ensures data portability and user control. Humans interact via a dedicated operator UI or integrated "surfaces" like Slack, Teams, and email, while AI agents leverage CLI and SDKs to read and write the same pod state. This unified approach contrasts with fragmented chat-based AI interactions, providing a durable and accessible system for complex, long-running tasks.

Quick Start & Requirements

• Installation: A single-line installer script (curl ... | bash for Linux/macOS, PowerShell for Windows) sets up the full stack using Docker or Podman. A macOS desktop app is also available.
• Prerequisites: Docker or Podman is essential. LLM provider configuration (Anthropic-compatible or OpenAI-compatible API keys/endpoints) is required for agent functionality.
• Documentation: Official docs and quickstart guides are available at lemma.work.

Highlighted Details

• Unified Pod Architecture: Consolidates tables, files, agents, workflows, and approvals into a single, portable workspace.
• Broad Surface Integration: Connects agents and workflows to users via Slack, Microsoft Teams, Gmail, Outlook, Telegram, and WhatsApp.
• Flexible LLM Integration: Supports any Anthropic-compatible or OpenAI-compatible LLM endpoint, allowing use of existing keys and models without vendor lock-in.
• Local-First Operation: The full stack runs self-contained on the user's machine, with data remaining local unless explicitly shared.

Maintenance & Community

Hosted on GitHub under the lemma-work organization. Active CI is indicated by build status badges. Specific community channels (e.g., Discord, Slack) or detailed contributor information are not detailed in the provided README excerpt.

Licensing & Compatibility

The platform employs a dual-licensing strategy:

• AGPLv3: Applies to the core backend (lemma-backend/) and frontend (lemma-frontend/), requiring network-delivered modifications to be open-sourced under the same terms.
• Apache-2.0: Covers client-side developer tools, including SDKs, CLI, and installer (agentbox/, lemma-stack/, lemma-cli/, lemma-python/, lemma-typescript/, lemma-skills/), permitting broader integration. Commercial licensing exceptions are available from Lemma for organizations facing AGPLv3 procurement constraints. Trademark rights are separate from software licenses.

Limitations & Caveats

The AGPLv3 license for core components may present adoption challenges for commercial SaaS offerings without securing a specific commercial license. Users must configure their chosen LLM provider credentials for agent functionality.