Connect Claude Code to EClaw for Autonomous Kanban Task Execution

What It Does

claude-code-eclaw-channel is an open-source bridge that connects Claude Code to EClaw's Kanban system. Instead of manually prompting Claude Code in your terminal, you can now assign it tasks directly on a Kanban board. When a card is assigned to Claude Code, the bridge:

1. Receives a webhook from EClaw
2. Routes the task to Claude Code via a fakechat plugin
3. Lets Claude Code execute autonomously (edit files, run tests, open PRs)
4. Reports progress back to the Kanban card via comments
5. Moves the card to "done" when complete

This creates a true A2A (AI-to-AI) workflow where EClaw entities can assign tasks to Claude Code as if it were a human teammate.

Setup

Getting started takes about 5 minutes:

git clone https://github.com/HankHuang0516/claude-code-eclaw-channel
cd claude-code-eclaw-channel
bun install
./setup-macos-permissions.sh  # macOS only

You'll need:

• Bun installed (the project uses Bun runtime)
• An EClaw account (free tier available)
• Claude Code installed and configured

After cloning, you'll configure the bridge with your EClaw API credentials and set up the webhook endpoint. The repository includes clear configuration examples.

Key Benefits for Claude Code Users

1. Phone-Based Permission System

The most significant improvement over standard Claude Code workflows is the permission system. Instead of using --dangerously-skip-permissions or manually approving every action in your terminal, permission requests route to your phone via EClaw's mobile app. You get push notifications and can approve/reject actions with a tap.

2. Zero Invasive Changes

The bridge doesn't modify Claude Code itself. It uses the existing fakechat plugin interface, which means:

• No breaking changes to your Claude Code installation
• Compatibility with future Claude Code updates
• You can still use Claude Code normally in your terminal

3. Progress Tracking and Audit Trail

Every action Claude Code takes gets logged as comments on the Kanban card. This creates a complete audit trail:

• What commands were executed
• What files were modified
• When tasks started and completed
• Any errors encountered

When To Use It

This bridge shines in specific scenarios:

Batch Processing Tasks

Instead of manually prompting Claude Code for each item in a list, create Kanban cards for:

• Refactoring multiple files
• Writing test suites for several components
• Updating dependencies across projects
• Generating documentation for API endpoints

Team Collaboration

If you're using EClaw with other team members (human or AI), you can now:

• Assign technical tasks directly to Claude Code
• See Claude Code's work alongside human teammates
• Use the same review process for AI-generated code

Scheduled/Deferred Work

Create Kanban cards for tasks you want Claude Code to handle later, then let the system process them when you're not at your computer.

Limitations to Consider

• The bridge requires EClaw (another service/tool to manage)
• Currently Bun-only (Node.js support may come later)
• Webhook setup requires a publicly accessible endpoint or ngrok/tunnel
• Phone permissions mean you need your device nearby for approval

The Technical Flow

Here's what happens under the hood:

EClaw Kanban → Webhook → bridge.ts → fakechat plugin → Claude Code
    Claude Code edits files, runs tests, opens PRs
        → POST /api/mission/card/:id/comment (progress)
        → POST /api/mission/card/:id/move (done)

The bridge acts as middleware, translating Kanban cards into prompts Claude Code understands, then translating Claude Code's output back into Kanban updates.

Getting Started Today

1. Clone and install the bridge repository
2. Create an EClaw account if you don't have one
3. Set up a webhook in EClaw pointing to your bridge
4. Configure permissions in the EClaw mobile app
5. Create a test card assigned to Claude Code

Start with simple tasks like "Create a README for project X" or "Run tests and report coverage" to see the flow in action.

gentic.news Analysis

This development represents a significant step toward integrating Claude Code into existing development workflows rather than treating it as a separate tool. The EClaw bridge follows a growing trend of MCP (Model Context Protocol) servers and integrations that extend Claude Code's capabilities beyond the terminal.

What makes this approach particularly interesting is how it addresses the permission problem that has plagued autonomous AI coding assistants. By routing permissions through a mobile app, developers get the safety of manual approval without the friction of constant terminal interruptions. This could become a model for other integrations seeking to make Claude Code more autonomous while maintaining control.

For teams already using Kanban methodologies, this bridge reduces the cognitive load of managing AI assistance. Instead of context-switching between your project management tool and terminal, Claude Code becomes another resource you can assign work to directly from your existing workflow. As more teams adopt AI coding assistants, integrations like this will be crucial for scaling their use beyond individual developer experimentation.

Resources

• GitHub Repository: https://github.com/HankHuang0516/claude-code-eclaw-channel
• EClaw: https://eclawbot.com
• Claude Code Documentation: https://docs.anthropic.com/claude/code