Shared-terminal MCP server for visible AI-assisted development
ripple, developed by Yotsuda, is an MCP server that creates a shared terminal for collaborative AI and human interaction in a single live console. The app presents real-time command visibility, interactive prompt handling, session persistence, and multi-shell management so an AI agent and a developer can operate together in one session. Key elements include support for bash, PowerShell, and cmd, automatic current-working-directory handoff, and ConPTY-based emulation. It targets developers and DevOps engineers who need transparent, interactive terminal workflows.
What tasks the app actually supports in terminal workflows
The app turns the terminal into a collaborative workspace by exposing AI actions in the same session as the user, which makes it suitable for command-line development, DevOps tasks, and interactive CLI tools that require back-and-forth input. Ripple supports interactive prompts and preserves session state, so tasks that depend on stepwise interaction or long-lived shells fit this model better than one-shot command dispatch.
How transparent and observable AI actions are during use
Real-time command visibility makes AI activity directly inspectable, removing silent execution from the workflow. The app displays AI-generated commands as they happen and maintains session persistence, which creates an auditable stream of terminal interaction. That behavior reduces the opacity of automated runs and helps developers verify each step while the AI operates in the live console.
Which environments and inputs the app accepts
The app supports major shells and host integrations, so it matches common developer environments. Supported shells include:
bash
PowerShell (pwsh)
Windows Command Prompt (cmd)
The server works with any host that implements the Model Context Protocol, for example Claude Desktop, and uses C# with ConPTY for high-fidelity terminal emulation.
How it fits into existing developer workflows and integration needs
Session persistence and automatic CWD handoff align with multi-step workflows, enabling the tool to keep context across switches between shells and tasks. Integration depends on an MCP-compatible host application, so adoption requires that host layer. Teams that need interactive debugging, package installs, or live DevOps commands benefit most from this model, while purely isolated automation pipelines may not gain the same advantages.
Practical choice for visible, interactive AI in the shell
The app is a practical option for developers and DevOps engineers who require live, reviewable AI interactions inside their terminal sessions. It improves transparency and supports interactive CLIs, but it suits workflows that accept a shared-session model rather than strict separation between agent and environment. For teams that prioritize inspectable, stepwise AI assistance, it represents a usable integration point into terminal-centered work.
Pros
Shared console shows AI-generated commands in real time
Supports bash, PowerShell (pwsh), and Windows cmd shells
Session persistence keeps state across multiple interactions
Handles interactive CLI prompts that break one-shot integrations
Cons
Requires an MCP-compatible host application to operate
Shared-session model may not suit strict separation or sandboxing needs
Built with ConPTY-based emulation, implying specific terminal emulation choices
Laws concerning the use of this software vary from country to country. We do not encourage or condone the use of this program if it is in violation of these laws. Softonic may receive a referral fee if you click or buy any of the products featured here.
