feat(vcs): jj driver, detection and process wrapper

[THEjacob1000]

What Changed

Bottom of a 4 PR stack for colocated jj: this, then #11382, #11383, #11384. A JjVcsDriver behind the existing VcsDriverRegistry, detected before git when the nearest marker is .jj, plus the JjProcess wrapper, availability check (floor jj 0.42.0), repo path resolution and a vcs.kind field on local status. CI gets a setup-jj action.

Why

The reopen Julius asked for on #2461, on the VcsDriver foundation from #2435 rather than a parallel subsystem like #743 and #1796. jj owns the working copy, the colocated .git stays the object store, so diffs and hosting don't fork. A jj repo with no reachable .git is reported unusable, not treated as git. Every jj call sets git.abandon-unreachable-commits=false so a status poll can't abandon commits.

Nothing routes to the driver yet. Closes #1126 as a stack.

Stack

1. This PR
2. feat(server): jj-native checkpoints and review diffs #11382
3. feat(server): route workflow, hosting glue, and discovery by vcs kind #11383
4. feat(web,mobile): jj terminology and shared vcs action logic #11384

Manual evidence

Manually exercised in the installed Linux desktop app, build 2a249f8e503a (the complete stack), using disposable repositories. These captures verify the integrated stack, not each lower PR in isolation.

Discovery/rescan reports Jujutsu available. A colocated repo gets bookmark controls; a non-colocated repo disables actions with the colocation explanation. Missing/old binaries, nested-repo precedence and config overrides are still untested.

Discovery and unsupported repository

Checklist

• This PR is small and focused
• I explained what changed and why
• I included before/after screenshots for any UI changes
• I included a video for animation/interaction changes

Evidence captured with Codex; test turns used GPT-6 Astra.

[omarshaarawi]

hey @THEjacob1000 , I'd love to help out on this to get it across the line but should you not wait on #2829 to merge before any work is done? unless your base is based on that branch then I only see this being throwaway work.

[THEjacob1000]

@omarshaarawi yeah there's like 4 PRs to get this over the line (so I'm not just merging a 12k line PR into an open source project). This one is the only one that doesn't conflict with that existing PR (the other 3 are stacked on top of 2829).

EDIT: Even though it wasn't going to conflict and could actually still be based on main, it makes more sense to have all 4 PRs in the same stack from the same branch (and WAY easier to track)

[cursor]

Bugbot is paused — on-demand spend limit reached

Bugbot uses usage-based billing for this team and has hit its on-demand spend limit.

A team admin can raise the spend limit in the Cursor dashboard, or wait for the next billing cycle to continue.

[THEjacob1000]

@juliusmarminge fresh take on jj against V2, following your close note on #2461. It's stacked on t3code/codex-turn-mapping (cumulative diffs coz GitHub won't stack from a fork) so it tracks V2 rather than waiting for the merge. Happy to retarget main once #2829 lands if it's preferable to be reviewed from there.

[t3dotgg]

Note

🤖 GPT-6 Astra responding on behalf of Theo

We're closing this as part of a cleanup of old PRs. Thank you for the time and care you put into this contribution.

The current comparison includes 1,477 files and 478,596 changed lines, with conflicts in 546 files against the target branch. Much of that comparison comes from diverged base history, beyond the change described here. The latest implementation commit is from 2026-09-20.

At this point, regenerating a focused patch against the current intended base would be easier to review and maintain than updating this PR. We appreciate the problem report and the work already done; this thread will remain available as a reference. If you revisit it, please confirm the current direction with us and link a fresh, focused PR back here.