A desktop app for isolated, parallel agentic development (opens in new tab)

I have a local application I developed that works extremely well for this. I.e. every thread tied to a repo creates it's own worktree, then makes it edits locally, and then I sync back to main. When conflicts occur, they are either resolved automatically if possible (i.e. another worktree merged into main first, those changes are kept so long as they don't conflict, if conflicted we get the opportunity to resolve, etc.).

At any merge-into-main from a worktree, the "non-touched" files in the worktree are automatically re-synced to main, thus updating the worktree with any other changes from any other worktree that have been already pushed to main.

Of course, multiple branches can also be used and then eventually merged into a single branch later..

---

Also, this is very clearly exactly the same thing OP does in their system, as per the README on their github link..

We chose not to use separate git worktrees under the hood for this functionality. Let me try to break down why, maybe there's an opportunity for me to learn more here.

In my head I separate between use cases of 1) "different tasks" and 2) "best of n, same task".

The app that we built already had the ability to separate changes into branches while in the worktree (on disk) it renders the integration of the branches. Our canonical use case back in the days was "A developer works on a feature branch and wishes to commit & publish a bugfix from separate branch". When we learned that people were using this for running multiple parallel agents we added some additional tooling for it.

So in practice what happens when you have multiple agents coding in parallel with GitButler is that the system captures information after an agent completes an edit (via the agent hooks) and uses that to 1) stage the particular edit to a branch dedicated to the agent and 2) perform a commit into that branch (GB can have multiple staging areas, one per applied branch).

The system will not allow multiple agents to edit the same file at the same time (via a locking mechanism in the pre-edit hook), but agents do see each others changes.

In the context of the "different tasks for different agents" use case, we have found that them seeing edits by others to have a positive effect on the outcomes. The first one that comes to mind is - no merge conflicts. But beyond merge conflicts, we have found that there is a lower likelihood of reaching a state where code diverges semantically.

In my own usage, I have found it helpful when I am hands on programming on something and wish to have an agent do some auxiliary task, for us to share a workspace (so that I can nudge it one way or another).

Is there something I am missing here? Of course for best-of-n of the same task this doesn't exactly make sense, but with regards to different tasks, what are some additional reasons to require full isolation? (as different worktrees would provide)

It's feasible, and it makes more sense than separating and merging later (or keeping patches in memory then applying in bulk).

Why do I say it's feasible? We have the technology, right? The IDE knows which file the user has in focus, and can orient agents to use tooling that would inform them of that fact when they're running. Similarly, that same tooling could just spend a little bit of time planning focus to spread to multiple agents in a way they won't overlap.

Maybe big repos, monorepos and so on are a limitation. If we were on the previous "small-to-middle interlinked projects" era, that division would come in naturally. You only really need multiple agents in parallel on a single project if that thing is big enough to have more than one angle to work on. It's a push-and-pull that changes with the times, maybe we're heading to a more granular way of doing things.

My suggestion would be to keep things simple, pragmatic, and save development time. While git worktrees are not perfect, and they require extra space, it is easier for people to understand and easily locate this kind of structure while also being able to execute commands in a somewhat isolated environment. I would be happy if the app simply automates this for me and creates a new worktree, branch, and agent at that location as soon as I click 'Add agent'.

Then the only issues become merge conflicts between different branches... this is where extra dev time could be allocated, and implement agents that automatically merge branches.

Docker? I typically want other kinds of isolation for services and things anyway - it’s got its own file system, you could have N versions on the same branch working without conflict (at least the conflict you’d have with work trees).

There is some more plumbing involved but…not much?

Edit - task boards are my first thought for the comms side for agents sharing info.

Then the changes can't be tested to even verify that they pass the compiler/linter, much less tested to confirm they actually work. The only way to fix this is to "modify local files", where "local files" is either a separate worktree (that you don't manage or need to know the location of, possibly even on a separate machine); or a hacky, vibe-coded, in-memory VFS; or somewhere in between.

It's annoying and hilarious at the same time.

1. https://github.com/brainless/nocodo

I started building it out that way but found it very challenging to create parity between the models. E.g. they have different tools, system prompts, interruption semantics, cost tracking etc. The spirit of the product is decoupling the LLM from the UI, so we went with a custom loop / tools that can perform decently across all models.