openlegion

OpenLegion provides a framework for building and running autonomous AI agents that can perform a wide range of computer tasks. You describe the type of team you need in plain language, and the platform creates a fleet of agents that operate in isolated Docker containers. The service is positioned for production use, with a focus on security and cost control.

The workflow starts with a single prompt that defines the team's role, sales desk, content studio, research group, or a custom combination. OpenLegion then provisions each agent, assigns a specific LLM, tools, and a budget, and isolates the agent in its own container. Agents communicate through a fleet model that uses a blackboard and pub/sub system, avoiding a single point‑of‑failure “CEO” agent. Because each container has its own credential vault, API keys and private keys never appear inside the runtime environment; a proxy handles all external calls.

The platform targets teams that need continuous automation without writing code. Templates are available for common functions such as lead research and outreach, code planning and PR review, or content creation from research to draft. You can also define custom roles, tools, and spending limits, allowing agents to browse websites, manipulate files, interact with wallets, or execute on‑chain transactions. Budgets are set per agent, with daily and monthly caps that automatically stop the agent when reached, helping prevent unexpected charges. A single dashboard lets you monitor the entire fleet, see spending, and adjust permissions as needed.

OpenLegion supports more than zero LLM providers, runs agents 24/7, and adds a zero‑markup layer on top of LLM usage. The service includes a 7‑day money‑back guarantee and early‑customer pricing that is locked in for the first users. Integration points mentioned are Discord and GitHub, and the platform advertises a “no‑coding required” experience for non‑technical users.

Key features

• Describe a team in plain English; the platform builds the agent fleet automatically
• Each agent runs in an isolated Docker container with its own budget and tool permissions
• Credential vault and proxy keep API keys and private keys out of containers
• Per‑agent daily and monthly spend limits with automatic hard cut‑off
• Built‑in templates for sales, content, research, and engineering use cases
• 24/7 operation with no markup on LLM usage