Skip to content
AI Tool

QM Review

QM is an open-source, multiplayer agent harness developed by Y Combinator that provides teams with collaborative, persistent, cloud-based environments for running AI agents.

shipped Aug 13, 2026freemium
QM — product screenshot

Why it matters

1Open-sourced under an MIT license in July/August 2026, generating over 2 million views on its announcement thread.
2Achieved more than 13,000 GitHub stars and a Hacker News thread exceeding 600 points.
3Supports multiple AI models including Pi, OpenCode, Codex, and Claude Code.
4Designed for startups and mid-sized companies (10-500 people) with at least one individual comfortable operating infrastructure.

About QM

Business Model
Open Source
Target Audience
Startups and Y Combinator teams
GitHubOpen Source

overview

What is QM?

QM is a multiplayer AI agent harness tool developed by Y Combinator that enables teams within organizations to run AI agents in collaborative, persistent, cloud-based environments. It provides a control plane for AI agents, allowing every person and project to have its own dedicated AI agent with independent memory, session history, files, artifacts, credential grants, permissions, and audience.

features

Key Features of QM

QM (Quartermaster) offers a robust set of features designed for collaborative AI agent deployment and management within organizational settings. Its architecture prioritizes security, flexibility, and integration with existing workflows.

  • Open-source under an MIT license, allowing for self-hosting and customization.
  • Multiplayer agent harness providing dedicated, contextualized AI agents for each person and project.
  • Deep integration with Slack and an optional web UI for seamless operation within existing communication channels.
  • Contextual Identity and Access Control managed by a headless TypeScript core, ensuring secure operations with audit capabilities.
  • Swappable AI Model Harness supporting various underlying models such as Pi, OpenCode, Codex, and Claude Code.
  • Isolated Sandbox Execution for agents, running commands within a dedicated sandbox per scope to enhance security and prevent unauthorized access.
  • Collaborative, persistent, cloud-based environments for running AI agents.
  • Flexible management of agents with easy administration.
  • Designed for work-related tasks, including internal notes, email, documents, and code repositories.

use cases

Who Should Use QM?

QM is specifically designed for organizations seeking to integrate AI agents into their daily operations for enhanced productivity and automation. Its multiplayer and collaborative nature makes it suitable for teams requiring shared AI capabilities.

  • Startups and Mid-sized companies (10-500 people): Ideal for organizations with at least one individual comfortable operating infrastructure, looking to leverage AI for various departmental tasks.
  • Teams within organizations: Enables collaborative AI in Slack and on the web for functions like accounting, legal, events, and engineering.
  • Employees: Supports tasks such as searching internal notes, email, documents, databases, and the web together.
  • Platform engineers: Facilitates working in existing code repositories, running tests, opening PRs, monitoring CI, and checking system logs.
  • Operations teams: Useful for collecting incident context, pulling logs, correlating information, and suggesting fixes with read-only scopes.

how to use

How to Use QM

QM functions as a control plane for AI agents, integrating into existing platforms like Slack and offering a web UI. Deployment involves setting up the open-source software within an organization's cloud infrastructure.

  • 1Deploy QM: Install the open-source QM software within your organization's cloud environment.
  • 2Integrate with Slack: Connect QM to your Slack workspace to enable agent interaction within channels.
  • 3Configure Agents: Set up dedicated AI agents for individuals or projects, defining their memory, permissions, and access to tools.
  • 4Define Workflows: Utilize agents for specific tasks such as triaging inboxes, managing code repositories, or tracking projects.
  • 5Administer QM: Use the admin API for tasks like granting permissions, managing model allowlists, and setting budgets.
  • 6Monitor and Audit: Leverage QM's audit capabilities to track agent activities and ensure compliance.

pricing

QM Pricing & Plans

QM is released under an MIT license, making the core software free to use. Organizations deploying QM will incur costs primarily related to cloud hosting, model tokens, and operational overhead.

  • Freemium: The QM software itself is free under an MIT license. Operational costs include cloud hosting, model tokens, platform time (deployment, identity management, backups, upgrades, incident response), cloud resources (core services, PostgreSQL, sandbox compute, storage), security work, and workflow engineering.

Pros

  • +Open-source under an MIT license, offering flexibility and cost control for self-hosting.
  • +Designed for collaborative, multiplayer use across an entire organization, providing contextualized agents for each person and project.
  • +Deep integration with Slack and an optional web UI, fitting into existing communication workflows.
  • +Supports a swappable AI model harness, allowing organizations to choose and switch between models like Pi, OpenCode, Codex, and Claude Code.
  • +Robust security features including a headless TypeScript core for identity, policy, scheduling, memory, and audit, with isolated sandbox execution.
  • +Proven internal use at Y Combinator, demonstrating significant productivity gains for a small team.

Cons

  • Requires at least one individual comfortable operating infrastructure for deployment and maintenance.
  • While the software is free, organizations incur costs for cloud hosting, model tokens, and operational overhead.
  • API is not available, which may limit custom integrations beyond its core Slack/web functionality.
  • Function calling is proprietary, potentially limiting customization or interoperability with external tools.
  • Primarily targets startups and mid-sized companies, potentially less suited for very small teams or large enterprises with different infrastructure needs.
  • Multimodality is limited to text, not supporting other data types like images or audio.

Similar Tools

QM vs Competitors

QM differentiates itself in the AI agent harness landscape through its multiplayer design, deep integration with existing enterprise tools like Slack, and its focus on providing contextualized agents for an entire organization rather than individual users.

1

Specializes in orchestrating role-playing autonomous AI agents into collaborative teams with defined roles and shared goals.

While both are open-source and support multi-agent collaboration, CrewAI focuses on a Python framework for building agent teams and workflows, whereas QM provides a broader 'harness' for running agents in persistent, cloud-based environments with Slack/web integration.

2

Provides an open-source autonomous agent framework with a GUI control room for end-to-end goal execution.

SuperAGI is designed for autonomous execution with a focus on minimizing human involvement, while QM emphasizes a multiplayer, collaborative environment for teams with human oversight and control.

3

An open-source framework for building conversational multi-agent systems where agents collaborate dynamically.

AutoGen provides a flexible framework for building multi-agent conversations and workflows, including a no-code UI (AutoGen Studio), while QM offers a ready-to-deploy, persistent, cloud-based harness for teams with built-in Slack/web integration.

4

An open-source framework for deploying multiple LLM-based agents in task-solving and simulation applications, with a focus on multi-agent collaboration and emergent behaviors.

AgentVerse is a framework for building and deploying multi-agent systems, emphasizing simulations and task-solving, whereas QM is a collaborative, persistent harness designed for teams to run and manage agents in a cloud-based environment.

More on Stork

Related AI Tools

Other tools in this category, matched by shared tags