EkyBot Lets Claude Code Talk to Other AI Agents via @mentions

What It Does

EkyBot is an open-source bridge that connects Claude Code with other AI agents—specifically OpenClaw (custom local agents) and Claude Cowork—into a single collaborative interface. Think of it as Slack for your AI agents: you create channels, @mention agents, and they work together on tasks while maintaining their individual runtimes.

For Claude Code users, this means your coding assistant can now directly ask other specialized agents for help. Need data analysis? @mention an OpenClaw agent running a data pipeline. Need documentation written? @mention Claude Cowork. All conversations stay in one place with full context.

Setup

Since EkyBot is open-source, setup involves:

1. Install the bridge from their GitHub repository
2. Configure your agents by connecting Claude Code (running locally via CLI), OpenClaw agents, and Claude Cowork
3. Create channels for different projects or workflows
4. Start collaborating using @mentions between agents

The key technical detail: Claude Code maintains its local runtime on your machine—it's not running in EkyBot's cloud. EkyBot handles the routing and conversation management while your actual Claude Code instance stays local and secure.

When To Use It

This shines for complex workflows where Claude Code needs specialized help:

• Data-intensive tasks: "@OpenClaw, query the production database for last month's user metrics and summarize them for me"
• Documentation generation: "I've refactored this API. @Claude Cowork, can you draft updated documentation based on the changes?"
• Multi-step deployments: Claude Code handles the code changes, then @mentions another agent to run tests, then another to deploy
• Research + implementation: Claude Cowork researches best practices for a feature, then @mentions Claude Code to implement it

The @mention syntax is natural: just type @AgentName followed by your request, and EkyBot routes it to the correct agent while maintaining conversation history.

Cost and Token Control

EkyBot includes a dashboard that tracks costs across all your agents. For Claude Code users on Pro/Max subscriptions, this gives visibility into:

• Token usage per agent and per channel
• Configurable daily/monthly budgets per agent
• Automatic memory compression to manage context windows
• A 4-level memory system (session, daily, long-term, project)

This is particularly valuable when combining Claude Code (subscription-based) with OpenClaw agents (API-billed models) — you see all costs in one place.

Security Considerations

Your Claude Code instance remains local. EkyBot creates an encrypted tunnel to route messages but doesn't host your Claude Code runtime. This maintains the security model you're used to with the Claude CLI while adding collaboration capabilities.

Coming Soon

The roadmap includes integration with n8n, LangChain, CrewAI, and "any agent with an API." This suggests Claude Code could eventually collaborate with automation workflows, complex agent chains, and custom tools.

gentic.news Analysis

This development follows the broader trend of AI agent specialization and collaboration we've been tracking. While tools like Cursor focus on integrating multiple AI capabilities into a single editor, EkyBot takes a different approach: connecting specialized agents that maintain their native environments. This aligns with our coverage of the growing "AI team" paradigm, where different AI systems handle different aspects of development workflows.

For Claude Code users, the immediate value is extending Claude's coding capabilities without leaving the collaborative interface. Instead of switching between Claude Code for development and other tools for research or data work, you can @mention the right agent for each task. This could significantly streamline workflows that involve both coding and adjacent tasks like data analysis, documentation, or research.

The open-source nature is notable—it contrasts with proprietary agent collaboration platforms and gives developers control over their agent infrastructure. As the ecosystem expands with promised integrations (LangChain, CrewAI), Claude Code could become the coding specialist in increasingly sophisticated AI teams.

Try It Now

If you regularly use Claude Code alongside other AI tools or custom agents:

1. Check the EkyBot GitHub for installation instructions
2. Set up a test channel with Claude Code and one other agent (Claude Cowork is the easiest start)
3. Try a simple workflow: Have Claude Cowork research a topic, then @mention Claude Code to implement a proof of concept
4. Monitor the cost dashboard to understand token usage patterns across agents

The key advantage for developers is eliminating context switching between different AI tools while maintaining each tool's specialized capabilities.