Our take
Newless than six months old and changing fast- Activity
- 76
- Community
- 50
- Issues
- 64
- Pull requests
- 64
+34 stars in the last 30 days
- Merges more pull requests than 92% of the projects we track
OpenCompany is a fast-moving, actively maintained experiment in running a configurable business through coordinated agents, but its own documentation cautions that it is not production-ready.
Good fit if
- You are a solo operator who wants to experiment with delegating business functions to agents while retaining key decisions.
- You can work with Rust-based software and provide a TinyHumans API key for agent work.
- You want to study or customize a company runtime whose core coordination uses TinyHiveMind and OpenHuman.
Look elsewhere if
- You need a stable production system, since APIs, CLI behavior, examples, and documentation may change without notice.
- Your deployment cannot use the TinyHumans service for agent capabilities, or you need agents to do real work without its API key.
- Your project cannot accommodate the obligations of the GPL-3.0 license.
All health signals
| Last commit | 2026-10-09 (10 days ago) |
|---|---|
| Commits, last 90 days | 100+ |
| Releases, last 12 months | 12 (latest v0.3.0, 2026-09-29) |
| Contributors | 16 (top contributor: 62% of commits) |
| Issues closed, last 90 days | 12+ (typically closed in 3 days) |
| Pull requests merged, last 90 days | 89 (typically merged in 1 day) |
| Project age | 3 months |
Checked on 2026-10-20 with the GitHub API.
Project history
Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.
Stars
339Forks
63Change since Nov 17, 2026: +4 stars·0 forks
Overview
OpenCompany is a Rust company runtime for one-person businesses, built on TinyHiveMind from TinyHumans AI and hosted over the OpenHuman agent runtime. It addresses the coordination gap between a single operator and the many functions needed to run a business: rather than handing tasks through a simple agent pipeline, its agents work together in desk-based hives, with the operator retaining important judgment calls.
The project offers a desktop app for macOS, as well as a host that can be run with Docker or Podman or built from source. Its company setups define agent roles, responsibilities, and human approval points, with examples covering businesses from software teams to research labs.
Key Features
- Runs specialist agents in shared desk hives, where they can message colleagues and work concurrently.
- Records conversation and episode activity, and can rebuild the record from a journal after a reload.
- Defines company rosters, agent responsibilities, and human sign-off points in
company.tomlfiles. - Includes example setups for 22 types of organizations, from venture studios to accounting firms.
- Supports a desktop app for macOS and self-hosted deployment paths, including Docker Compose.
- Provides a CLI check for company configuration and a Rust host with an Axum HTTP surface.
Use Cases
- A solo founder can delegate recurring research, marketing, software, or support tasks to specialized agent roles while keeping strategy and approvals.
- An operator can run a predefined company setup, such as a marketing agency or product team, instead of assembling an agent workflow from scratch.
- A developer can customize a company roster and its approval boundaries through configuration, then run it locally or on a self-hosted host.
- A team exploring coordinated agent behavior can inspect how desk-based hives, messaging, and bounded work episodes operate in a company runtime.
What you need
Detected in the repository
- Rust (from Cargo.toml)
- Automated checks on GitHub Actions
License in plain words
GPL-3.0strong copyleft
- Commercial use: yes
- Modify and redistribute: yes
- You must keep: the license notice
- Share your changes: yes, the full source of anything you distribute that includes it, under the GPL
- Includes an explicit patent grant from the contributors:
A summary, not legal advice: the LICENSE file is what applies.
Getting Started
Clone the repository with its submodules and launch the marketing demo:
git clone --recurse-submodules https://github.com/tinyhumansai/opencompany.git
cd opencompany
export TINYHUMANS_API_KEY="th-..."
./scripts/launch-demo.sh marketing up
Then open http://localhost:5173. See the README for setup details and other demo companies.
Alternatives
- Orkas: Orkas directs a Commander and specialists from a local desktop chat for research, documents, and coding, rather than running a human-led business as a shared agent hive.
- deer-flow: DeerFlow is a self-hosted harness for long-running research, coding, and content tasks, rather than a Rust runtime organized around a one-person business.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| opencompany | Rust | GPL-3.0 | 307 | New |
| Orkas | JavaScript | MIT | 2.1k | New |
| deer-flow | Python | MIT | 83.4k | Active |
Considerations
- The project is explicitly work in progress. Its measured history shows it is new and changing quickly, despite frequent recent commits and releases.
- It has 16 contributors and recent pull requests are being merged quickly; contributor activity is not concentrated in one person. Recent issue closures have also been relatively quick, though that does not establish production readiness.
- Agent work requires a TinyHumans API key. Self-hosting the host does not remove this dependency for agent capabilities.
- The README describes Docker or Podman as the simplest host setup and documents a macOS desktop app. Windows quickstart instructions require a POSIX-style terminal such as WSL or Git Bash.
- GPL-3.0 licensing may impose obligations on modified or redistributed versions; review the license before incorporating the project into a product.
Found this useful?
Share it with someone who would like opencompany.