imcodes
The IM for agents. Shared Agent Context & Memory, supervised execution, and cross-agent audit across AI providers.
The imcodes MCP server delivers a runtime-scoped tool surface for AI coding agents, enabling cross-provider memory, inter-agent communication, and supervised execution. It connects terminal-based agents like Claude Code, Codex, Gemini CLI, and Qwen to shared context, remote terminals accessible from mobile or web, and structured multi-agent collaboration workflows. The platform maintains agent work sessions across devices, injects relevant past work into new conversations, and orchestrates team-based code review before implementation. Self-hosted deployment keeps all agent sessions and memory on your own infrastructure.
Key Features
Use Cases
- 01Continue coding-agent workflows from mobile when away from your desk by accessing live terminals and file changes
- 02Reduce single-model blind spots by running cross-provider team audits before code lands
- 03Coordinate multi-agent workflows where one agent reviews changes while another runs tests
- 04Schedule recurring agent tasks like log monitoring, CI failure diagnosis, or webhook-triggered code reviews
- 05Operate a fleet of development machines from a single conversation using Controlled Nodes
- 06Preview local development servers on mobile devices without deploying or exposing public URLs
Related MCP Servers
View moremarkitdown
Python tool for converting files and office documents to Markdown.
firecrawl
The context API to search, scrape, and interact with the web at scale. 🔥
prompts.chat
f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.
langflow
Langflow is a powerful tool for building and deploying AI-powered agents and workflows.
imcodes — FAQ
What is the imcodes MCP server?+
The imcodes MCP server is a daemon-managed stdio MCP server that provides AI agents with runtime-scoped tools for memory recall, inter-agent messaging, and scheduled tasks. It works alongside a WebSocket server and daemon to deliver shared context, supervised execution, and remote terminal access across multiple AI coding agents.
How do I install and connect the imcodes MCP server?+
Install globally with npm install -g imcodes, then bind your machine to a server instance using imcodes bind with your API key URL. For self-hosting, run imcodes setup --domain with your domain name to deploy the full stack including PostgreSQL, server, and Caddy with automatic HTTPS.
Which AI clients work with imcodes?+
The imcodes MCP server supports SDK-backed providers including Claude Code, Codex, Gemini CLI, OpenClaw, Qwen, plus GitHub Copilot and Cursor via its broader platform features. Transport-backed agents receive the managed MCP tools automatically when connected.
Do I need API keys or prerequisites?+
You need Node.js 22 or higher, tmux on Linux/macOS (Windows uses built-in ConPTY), and at least one AI coding agent CLI or SDK installed. For self-hosting you also need Docker and a domain with DNS configured.
Is imcodes free and open source?+
Yes, imcodes is MIT-licensed open-source software. The shared instance at app.im.codes is for testing only; self-hosting on your own infrastructure is strongly recommended for production use.
How does imcodes handle memory and context across agents?+
It continuously converts completed agent work into reusable summaries stored locally, with optional cloud sync for cross-device access. Enterprise teams can publish memory into workspace scopes, and multilingual semantic search retrieves relevant past fixes across English, Chinese, Japanese, Korean, Spanish, and Russian codebases.
How do I install imcodes?+
Open the source repository on GitHub and follow its README. imcodes is a mcp server — MCP Agents Market links you directly to the official repo.
Is imcodes free?+
imcodes is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.