Our take
Activeregular commits and releases- Activity
- 87
- Community
- 52
- Issues
- 67
- Pull requests
- 70
+34 stars in the last 30 days
- Releases more often than 90% of the projects we track
CCCC is an active, frequently updated collaboration layer with substantial runtime breadth, best suited to teams that value persistent coordination over a lightweight single-agent workflow.
Good fit if
- You run multiple coding agents and need durable history, explicit delivery and reply tracking, or task handoffs.
- You want a local-first daemon with Web, CLI, MCP, and IM access, without operating a separate database or message broker.
- You can work within Apache-2.0 terms and the supported Linux, Apple Silicon macOS, or Windows targets.
Look elsewhere if
- You need a pure DAG workflow orchestrator; CCCC describes itself as a collaboration kernel, not a replacement for one.
- Your main need is a simple single-agent helper, where the project says its multi-participant coordination adds less value.
- You require every integration to have the same setup and interaction model; runtime capabilities and requirements vary.
All health signals
| Last commit | 2026-10-19 (1 day ago) |
|---|---|
| Commits, last 90 days | 100+ |
| Releases, last 12 months | 73 (latest v0.4.44, 2026-10-20) |
| Contributors | 12 (top contributor: 53% of commits) |
| Issues closed, last 90 days | 30 (typically closed in 5 days) |
| Pull requests merged, last 90 days | 30 (typically merged in 0 days) |
| Project age | 14 months |
Checked on 2026-10-21 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
1,313Forks
112Change since Nov 13, 2026: +4 stars·+1 forks
Overview
CCCC is a local-first coordinator that brings coding agents together as a persistent, coordinated team across runtimes and machines. Its Rust implementation keeps shared messages, tasks, and context in durable local state, addressing the coordination gaps that arise when agents need to hand off work, confirm delivery, or resume after restarts.
A single daemon provides a shared control plane for the Web UI, CLI, MCP tools, and IM bridges. CCCC supports Linux, Apple Silicon macOS, and Windows; it is aimed at people running multiple coding agents who want a durable collaboration layer rather than a workflow engine that executes a fixed DAG.
Key Features
- Append-only event history, with separate stores for tasks and shared context.
- Message routing and distinct stored, runtime-delivery, read, and reply states.
- Send, Send + Reply, and Mail modes, including an ordered, consuming Inbox for Mail.
- Mixed-runtime groups, including CLI agents, browser-based integrations, and custom command-line agents.
- Web UI, CLI, MCP, and supported IM bridges operating on daemon-owned state.
- Cross-instance collaboration through same-account discovery, approved Group connections, or direct connections between Groups.
- Workspace file tools, Git inspection, tiled terminals, and configurable automation rules.
Use Cases
- A developer coordinating Claude Code and Codex on a shared repository, with visible task ownership and delivery status.
- A team keeping long-running agent groups available for remote monitoring and coordination through the Web UI or an IM bridge.
- Operators connecting selected Groups across machines or trusted teams without sharing each instance's full state.
- Teams using multiple supported runtimes that want collaboration state and message history to persist independently of agent sessions.
What you need
Detected in the repository
- Python (from pyproject.toml)
- Rust 1.88 or newer (from Cargo.toml)
- A test suite and automated checks on GitHub Actions
License in plain words
Apache-2.0permissive
- Commercial use: yes
- Modify and redistribute: yes
- You must keep: the license, the NOTICE file and a note of your changes
- Share your changes: no
- Includes an explicit patent grant from the contributors:
A summary, not legal advice: the LICENSE file is what applies.
Getting Started
Install with the recommended website installer, then launch CCCC:
curl -fsSL https://chesterra.github.io/cccc/install.sh | sh
cccc
Open http://127.0.0.1:8848 for the local Web UI. See the README for Windows installation, agent setup, and security guidance before enabling remote access.
Alternatives
- codeg: Codeg centers coordination on Agent Client Protocol sessions, code changes, and tasks, rather than durable team messaging across runtimes and groups.
- claude-code-by-agents: Claude Code by Agents coordinates work through @mentions and threaded rooms backed by OpenAgents workspaces, rather than a Rust messaging coordinator with delivery tracking.
- coven: Coven focuses on local, project-bounded agent tools and shared session management, rather than coordinating a persistent team across machines and trusted groups.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| cccc | Rust | Apache-2.0 | 1.3k | Active |
| codeg | Rust | Apache-2.0 | 4k | Active |
| claude-code-by-agents | Swift | MIT | 892 | Quiet |
| coven | Rust | MIT | 73 | Active |
Considerations
- The project is actively maintained, with regular commits and releases, and recent issue and pull request handling. It has a contributor base rather than depending mostly on one person, but it is still a young project, so expect integrations and operational details to evolve.
- Agent runtimes generally need to be installed and authenticated separately. Supported integrations have differing setup steps, interaction surfaces, and capability boundaries.
- Remote Web access requires an Admin Access Token and an appropriately protected route. The daemon's local IPC is a trusted interface, not an authenticated public endpoint.
- Live voice features are described as experimental. Some browser-based model integrations also have explicit setup requirements and compatibility boundaries.
- The project requires Rust 1.88+ for source builds and Node.js 24; the recommended installer avoids requiring a Rust toolchain for end users. It does not require a separately operated database server, message broker, or Docker.
Found this useful?
Share it with someone who would like cccc.