Jinn: Run Claude Code as a Multi-Agent Team with Cron Jobs and Slack Integration
Open SourceScore: 87

Jinn: Run Claude Code as a Multi-Agent Team with Cron Jobs and Slack Integration

Jinn is an open-source gateway daemon that turns Claude Code CLI into a multi-agent system with scheduling, Slack integration, and a web dashboard.

3d ago·4 min read·7 views·via reddit_claude, hn_claude_code
Share:

Jinn: Run Claude Code as a Multi-Agent Team with Cron Jobs and Slack Integration

If you're using Claude Code CLI daily, you've probably hit its limitations: no background jobs, no team collaboration, no external integrations. Jinn solves this by building a thin orchestration layer on top of the CLI you already use.

What Jinn Does

Jinn doesn't replace Claude Code CLI—it enhances it. The key insight: Claude Code already handles tool use, file editing, memory (with Auto Memory), and multi-step reasoning. Jinn adds what's missing:

Multi-Agent Organization System
Define employees as YAML files with departments, ranks, and managers. Agents can delegate tasks to each other, creating hierarchical teams that work together on complex projects.

Cron Scheduling
Set up hot-reloadable background jobs that run automatically. Examples from the creator include daily standups, content pipelines, and inbox monitoring—all powered by Claude Code.

Slack Integration
Thread-aware message routing with reaction-based workflows. Claude Code agents can respond to Slack messages, creating seamless collaboration between your team and AI assistants.

Web Dashboard
Visual interface showing chat history, org structure, kanban boards, cost tracking, and cron job visualization. Monitor everything from one place.

Skills System
Markdown playbooks that Claude Code follows natively. Create reusable workflows and procedures that agents can execute consistently.

How It Works

Jinn wraps the Claude Code CLI (and optionally Codex) as engines. When you make a request through Jinn, it routes it to the appropriate Claude Code instance, handles the orchestration between multiple agents if needed, and manages the output.

Self-Modification Capability
Agents can edit their own config, skills, and org structure at runtime with hot reload. This means your AI team can evolve and optimize itself based on performance.

Max Plan Compatibility
Since Jinn delegates to the official Claude Code CLI, it works with any plan that supports CLI access—including Max plans. No special API keys or permissions needed beyond what you already have.

Installation & Setup

# Clone the repository
git clone https://github.com/yourusername/jinn.git
cd jinn

# Install dependencies
npm install  # or yarn install

# Configure your agents
cp config/agents.example.yaml config/agents.yaml
# Edit agents.yaml to define your organization

# Start Jinn
npm start

Basic Agent Configuration

agents:
  - name: "code-reviewer"
    department: "engineering"
    rank: "senior"
    manager: "cto"
    skills:
      - "code-review.md"
      - "testing.md"
    schedule:
      - "0 9 * * 1-5": "review-pull-requests"

When To Use Jinn

Team Projects - When you need multiple specialized Claude Code instances working together (e.g., one for frontend, one for backend, one for testing).

Automated Workflows - For recurring tasks like daily code reviews, dependency updates, or documentation generation.

Slack Integration - When you want Claude Code accessible to your entire team through Slack channels.

Complex Orchestration - For projects requiring sequential or parallel AI tasks with handoffs between different agents.

Limitations & Considerations

Jinn adds overhead—you're running additional processes alongside Claude Code. For simple one-off tasks, stick with direct CLI usage. The web dashboard requires a separate server, and Slack integration needs bot configuration.

Since Jinn builds on the existing CLI, it inherits Claude Code's limitations too: token limits, context windows, and any CLI-specific constraints.

The Future of Claude Code Orchestration

Jinn represents a growing trend: developers building orchestration layers on top of AI tools rather than waiting for native features. This approach lets you customize workflows to your specific needs while benefiting from Claude Code's continuous improvements.

The creator mentions this is just the beginning—expect more integrations (Discord, Teams, email) and advanced features like agent marketplaces where you can share pre-configured AI employees.

Try it today: The project is open-source and actively maintained. If you're already using Claude Code CLI for serious work, Jinn could transform how you collaborate with AI.

AI Analysis

Claude Code users should start thinking beyond single-session interactions. Jinn enables persistent AI workflows that run automatically and collaborate. Here's what to do differently: 1. **Define AI roles explicitly.** Instead of one Claude instance doing everything, create specialized agents in Jinn: a code reviewer, a documentation writer, a bug fixer. This matches how human teams work and improves results. 2. **Schedule repetitive tasks.** Any Claude Code workflow you run manually more than twice should be automated via Jinn's cron system. Daily code reviews, weekly dependency checks, or automated testing pipelines can run while you sleep. 3. **Use the skills system.** Document your best practices in markdown playbooks. When you discover a particularly effective prompt sequence for a task (like refactoring React components), save it as a skill file that any agent can use. 4. **Start with one integration.** Don't try to implement Slack, cron, and multi-agent all at once. Begin by wrapping your existing Claude Code CLI usage with Jinn, then add scheduling for one recurring task, then expand from there.
Original sourcereddit.com

Trending Now

More in Open Source

View all