github.devotd.
NewCoding agents

pawanosman/opencursor — an open-source coding agent for VS Code

pawanosman/opencursor is an open-source agent that runs coding tasks with a choice of models inside VS Code, and the latest verifiable release is 0.2.0.

Repositorypawanosman/opencursor ↗

pawanosman/opencursor is an open-source agent that runs coding tasks with a choice of models inside VS Code, and the latest verifiable release is 0.2.0.

GitHub description: Open-source Cursor-like AI coding agent for VS Code - agentic chat, multi-provider LLMs (OpenAI, Ollama, llama.cpp), semantic search, and MCP support

stars
★ 6K
Forks
1K
Language
TypeScript
License
MIT
Latest release
0.2.0
Read original source ↗View on GitHub ↗

pawanosman/opencursor is an open-source AI coding agent for VS Code. The latest verifiable release is 0.2.0. The provided record contains no changes for this version, so no added or altered feature can be confirmed. The focus therefore stays on the features and scope recorded in the repository rather than on the specifics of this update.

It is a tool that chats inside the editor, explores code, edits files, runs terminal commands, and then lets you review the changes. You can use your own API key or connect OpenAI, Ollama, or llama.cpp models, and pick the model and permissions to work with. Results can be reviewed and then kept or reverted. Semantic search, MCP support for connecting external tools, sub-agents, and a planning mode are described as well. It is MIT licensed and the primary language is TypeScript. The key point is that it is not locked to a single model, but provides a workflow that runs from model selection in the editor through change review.

That said, the provided evaluation gives a momentum score of 2.6, with no novelty, technology, or quality scores. The provided record contains neither a 0.2.0 changelog nor performance measurements. So it is hard to judge actual editing accuracy or pace of development from the feature list alone. What can be confirmed today is limited to the scope: an implementation that handles multiple models and an agent workflow inside VS Code.