Verified project record

PawanOsman/OpenCursor

OpenCursor is an open-source VS Code extension that adds a local-first AI coding agent capable of editing files, running commands, and searching code semantically. It supports fully offline operation via local models or connects to optional cloud providers.

Project overview

Project type: MCP · AI Agent · AI Coding
Use cases: Chat Assistants · Coding & Development
Deployment: Refer to project documentation
License: MIT

Best for

Developers using VS Code who need a local-first AI agent for editing files, running commands, and semantic search with optional external services or fully offline models.

Key capabilities

Built-in support for llama.cpp, Ollama, and local ONNX MiniLM embeddings allows the agent to run completely offline without internet.
Multi-tab sidebar chat with @-mentions for files, folders, git commits, and more.

Limitations and risks

Native runtime dependencies are downloaded once on first activation because they are too heavy to ship in the VSIX package. Telemetry practices are not documented.

Getting started

Install OpenCursor from the VS Code Marketplace, open the sidebar, pick a provider on first run, and start chatting.

Evidence and sources

README: The open-source AI coding agent for VS Code — built local-first. Chat with an agent that reads your workspace, edits files, runs commands, and searches your codebase semantically.…
README: Airplane-mode coding — local model + local index = a fully working AI agent, offline.
Release: 0.0.2
README: Ask questions in plain language — *"where do we refresh the auth token?"* — and the agent automatically finds code by *meaning*, not keywords.
README: OAuth sign-in — connect your existing **Claude Code**, **OpenAI Codex**, or **Google Antigravity** account and use your subscription's models directly in VS Code.

Reviewed at: 2026-07-16T15:54:40Z