Unify AI coding tools with a single sync layer across agents

AI-powered coding assistants have transformed how developers build software, but their fragmented ecosystems create workflow silos. Teams often juggle multiple tools—Claude Code for quick CLI tasks, Gemini CLI for deep context searches, and Cursor or Copilot as primary IDEs. However, a critical gap persists: these tools operate in isolation. Configurations set in one environment rarely propagate to another, forcing users to manually replicate prompts, system instructions, and MCP server setups across directories. This duplication wastes time and introduces inconsistency.

To bridge this divide, a new open-source project named Wasla—meaning "connection" in Arabic—has emerged as a universal synchronization layer for AI coding tools. Developed to streamline cross-platform agent management, Wasla eliminates redundant setup and ensures configurations remain consistent across all supported environments.

How Wasla Unifies Your AI Coding Tools

Wasla functions as both a CLI utility and a background daemon, designed to monitor and synchronize agent instructions and MCP servers across multiple AI platforms. It operates entirely locally, requires no cloud dependencies, and can be deployed in under a minute. The tool currently supports popular environments including Claude Code, Gemini CLI, Cursor, and others, with plans to expand compatibility.

Setting up Wasla is straightforward. For example, to integrate Gemini CLI into your workspace, users can run:

npx @untitled-devs/wasla setup gemini --scope workspace

This command configures the tool within the specified scope—either workspace-specific or user-wide—ensuring your agent definitions are immediately available wherever you work.

Behind the Scess: Smart Sync Without Duplication

Rather than copying entire configuration files—which risks version conflicts and storage bloat—Wasla employs a "Latest is Greatest" strategy centered on lightweight reference stubs. The system scans designated tool directories (such as ~/.claude/ and ~/.gemini/), then creates minimal pointers that reference the original files. When an AI tool loads a stub, it automatically resolves to the master version located in the originating directory.

Here’s how the directory mapping works in practice:

• ~/.gemini/agents/planner.md (Original file)
• ~/.claude/agents/planner.md (Stub that points to the Gemini original)
• ~/.openclaw/agents/planner.md (Stub that also points to the same original)

Whenever a user edits an agent instruction in one tool, Wasla designates that tool as the source of truth. The next time a sync occurs, all connected tools receive the updated configuration via their respective stubs. This approach minimizes storage overhead while preserving real-time consistency.

Key Features for Developer Productivity

Wasla introduces several features aimed at reducing configuration overhead and enhancing collaboration:

• Dual-Scope Synchronization
• Use --scope workspace to keep prompts and configurations tied to a specific project or Git repository.
• Opt for --scope user to maintain globally accessible agent definitions across your entire system.

• Daemon Watch Mode

Running wasla watch launches a silent background process that monitors file changes. It instantly propagates updates across all connected tools, ensuring your latest instructions are always in sync without manual intervention.

• Interactive Visual Dashboard

Execute wasla visualizer to launch a local React-based dashboard. The UI provides a clear overview of all connected agents, their instructions, and associated MCP servers, making it easy to audit and manage your AI toolchain.

Quick Start Guide for Developers

Getting started with Wasla is simple and requires minimal setup. Users can install the tool globally or run it directly via npx:

npx @untitled-devs/wasla watch --scope user

The project is fully open-source under the MIT license, encouraging community contributions and feedback. Developers are invited to share thoughts on architecture improvements, suggest new tool adapters, or report issues through the project’s public repository.

As AI coding assistants continue to evolve, tools like Wasla play a crucial role in bridging fragmented ecosystems. By centralizing configuration management and eliminating redundant setup, developers can focus on writing code rather than managing tools. The future of AI-assisted development may soon hinge not on isolated brilliance, but on seamless interoperability.