# Using vsrocq language server with Emacs

**URL:** <https://discourse.rocq-prover.org/t/using-vsrocq-language-server-with-emacs/3119>\
**Category:** Developing plugins\
**Tags:** coq-lsp\
**Created:** [September 29, 2026, 9:36am UTC](https://discourse.rocq-prover.org/t/using-vsrocq-language-server-with-emacs/3119 "2026-09-29T09:36:48Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Michalis](https://sea1.discourse-cdn.com/flex001/user_avatar/discourse.rocq-prover.org/michalis/32/1302_2.png) [@Michalis](https://discourse.rocq-prover.org/u/Michalis)\
**Post date:** [September 29, 2026, 9:36am UTC](https://discourse.rocq-prover.org/t/using-vsrocq-language-server-with-emacs/3119/1 "2026-09-29T09:36:48Z")

</div>

Has anyone looked at integrating the vsrocq language server with emacs? I have been using [jpoiret/rocq-mode.el: An Emacs mode for the Rocq theorem prover, using coq-lsp - Codeberg.org](https://codeberg.org/jpoiret/rocq-mode.el), and I really like its workflow over Proof-General, but I wonder if it would be better to base it on vsrocq, rather than rocq-lsp, since the former seems to be the more full-featured language server.

If nothing exists, I might experiment with either adapting rocq-mode or building something new. Curious if that is something people would be interested in.

---

<div class="post-metadata">

**Author:** ![TDiazT](https://sea1.discourse-cdn.com/flex001/user_avatar/discourse.rocq-prover.org/tdiazt/32/196_2.png) [@TDiazT](https://discourse.rocq-prover.org/u/TDiazT)\
**Post date:** [September 29, 2026, 10:13pm UTC](https://discourse.rocq-prover.org/t/using-vsrocq-language-server-with-emacs/3119/2 "2026-09-29T22:13:33Z")

</div>

Hi Michalis, yes, I will be working on VsRocq soon and one of my intentions is to integrate the language server with Josselin’s project. Feel free to reach out in Zulip (or privately here if you prefer) and maybe we can organize some meeting with Josselin and the other developers of VsRocq, to setup a route to have this working 🙂.

---

<div class="post-metadata">

**Author:** ![colorfulcms](https://avatars.discourse-cdn.com/v4/letter/c/8491ac/32.png) [@colorfulcms](https://discourse.rocq-prover.org/u/colorfulcms)\
**Post date:** [October 27, 2026, 3:06pm UTC](https://discourse.rocq-prover.org/t/using-vsrocq-language-server-with-emacs/3119/3 "2026-10-27T15:06:23Z")

</div>

As someone who’s only just started looking at Rocq, I’d be interested in this too. From a newcomer’s perspective, a lot of the pain is that each editor ends up with its own slightly different setup, so having vsrocq as a shared backend that Emacs can also talk to would make documentation and troubleshooting much easier. It might be worth writing down which LSP features rocq-mode currently depends on, so it’s clear what a vsrocq-based version would need to support.

---

<div class="post-metadata">

**Author:** ![michaelwilson](https://avatars.discourse-cdn.com/v4/letter/m/6bbea6/32.png) [@michaelwilson](https://discourse.rocq-prover.org/u/michaelwilson)\
**Post date:** [November 13, 2026, 9:14pm UTC](https://discourse.rocq-prover.org/t/using-vsrocq-language-server-with-emacs/3119/4 "2026-11-13T21:14:08Z")

</div>

The feature inventory colorfulcms suggests seems like the right first step. I’d split it into standard LSP capabilities and any VsRocq-specific requests or notifications: the former might be testable quickly by pointing eglot at VsRocq on a small file, while the latter will determine how much rocq-mode needs to change. That baseline could also make the Zulip/meeting discussion more concrete.
