DevSpace is a self-hosted MCP server for developers who want ChatGPT or Claude to work directly in projects on their own machine. It gives a connected assistant access to selected folders and the local terminal, so it can make code changes and check them against the project's actual tools.
The server runs locally. ChatGPT connects over the internet through an HTTPS tunnel you control, with password approval for the connection. Project files and command execution stay on your machine, while the connected AI service can read workspace content through the tools you expose. You choose which project folders it may open; access within a workspace includes shell execution.
An assistant can read and edit files, search a repository, and run tests, builds or Git commands. Isolated Git worktrees let parallel coding sessions work in separate checkouts. DevSpace can also delegate tasks to coding agents you select as subagents, and you can use it from coding agents without a ChatGPT connection.
It follows project guidance in AGENTS.md and CLAUDE.md and discovers skills in local skill folders. In ChatGPT Apps-compatible hosts, tool cards and optional change summaries show what the assistant did.
DevSpace is open source under the MIT license. It runs on Linux and macOS, plus Windows with a Bash-compatible environment such as Git Bash or WSL. Its local runtime requires Node.js, npm, Git and Bash; Windows with only PowerShell or cmd.exe isn't supported.
Claim this page and we'll verify you by hand. DevSpace gets the verified badge, and you can upgrade the listing to be featured on localhosted. Proud to be listed? Put our badge on your site.
Want more people to find DevSpace?Promote it
Something wrong or outdated on this page?
1.5KUpdated 3 weeks agoMIT
macOS · Windows · Linux · Web#Context compression#Human approval#MCP
agentchattr is a local chat server for developers who want multiple AI coding agents to work together in one conversation. Humans and agents can mention each other to prompt a response, so you don't have to carry messages between separate terminals. It runs on Windows, macOS and Linux, with a browser interface.
45.6KUpdated 4 hours agoMIT
macOS · Windows · Linux · Web#Git integration#Knowledge graphs#MCP
Codebase Memory MCP gives AI coding agents a persistent map of a repository's structure, so they can answer questions about dependencies and call chains without repeatedly reading whole files. It's for developers using Claude Code, Codex CLI, Cursor or other MCP-compatible clients who want their agents to spend less context on code exploration.
1.4KUpdated 1 month agoApache-2.0
macOS · Windows · Linux#Batch processing#Code execution#Human approval
1KUpdated 19 hours ago
macOS · Windows · Linux#Agent Skills#Git integration#MCP
Reticle lets AI coding agents verify changes against a running web or desktop app. It's for developers who want evidence that a fix works, including failures a screen can hide: duplicate payments, failed requests or application state that disagrees with the interface. It runs locally on macOS, Windows and Linux and connects to MCP agents such as Claude Code, Cursor and Codex CLI.
3.7KUpdated 2 weeks agoApache-2.0
macOS · Windows · Linux#MCP#OpenAI-compatible API#Streaming inference
1KUpdated 4 weeks agoApache-2.0
macOS · Windows · Linux#Knowledge graphs#MCP#Persistent memory
FastCtx gives coding agents structured access to local repositories through MCP. It's an open source Rust runtime for developers who want agents to spend less context on shell syntax and parsing terminal output. It runs on Windows, macOS and Linux under the Apache 2.0 license, with dedicated connections for ChatGPT App and Codex CLI and support for other MCP clients.
Anything Analyzer is a desktop protocol analyzer that brings browser traffic and HTTPS proxy captures into the same session for AI analysis. It runs on Windows, macOS (Intel and Apple Silicon) and Linux. It's for developers and security researchers investigating APIs, authentication flows, request signatures and JavaScript encryption.
GrapeRoot is a local MCP tool for developers who want their AI coding assistant to understand a project's structure without repeatedly searching through files. It builds a graph of the codebase and supplies relevant code before each prompt reaches the assistant. It runs on macOS, Linux and Windows.