@jenseralmeida this is close to what I wanted too, so I put together a workaround: the agent writes the plan as an artifact that lives outside the repo (under ~/.ai-artifacts), and I select any text, comment on it inline and send all the comments back in one go. I built an extension for it (disclosure: I’m the author) and wrote up how it works and where it falls short here: [Making Cursor’s agent write plans as artifacts so I can comment on them inline].