eonist

Here's a comprehensive implementation plan for AI Rebase, drawing from issue #177 , the existing split architecture in issue #196 , and the codebase structure .

High-Level Architecture

The AI Rebase is the inverse of the split feature — instead of one commit → many, it's many commits → fewer logical ones. You can mirror the existing split pipeline but reverse the data flow. The key files to model after are omni_commit.rs on the Rust side and the commit view components in src/views/commit/ .

Phase 1: Data Collection (Rust Backend)

Add a new Tauri command in src-tauri/src/ (e.g., omni_rebase.rs) that:

eonist

Alright André, let's put Ashera AI through the gauntlet of your own Substack playbook. This is going to sting. 🔥

The Tagline: "GTM, Run by AI"

Your article says to "craft a crystal-clear value proposition that hooks in 5 seconds" and lead with a "crisp tagline". Ashera's tagline is "GTM, Run by AI". That's three letters (GTM), a comma, and a buzzword. It tells me absolutely nothing about what this product actually does. Is it a CRM? A dialer? A sales coach? A pipeline forecaster? A person skimming Product Hunt has already scrolled past before their brain finishes parsing the acronym. This is textbook "vague product positioning" — the #1 item on your own "Will Not Get You Featured" list. reddit

Me-Too Clone Alert 🚨

Your article explicitly warns: "Search Product Hunt for similar products first. If 10+ clones already exist, you're too late". AI-powered sales call analysis? Let's see… Go

eonist

AI Rebase

We can also do the reverse. With code agents often having a tendency to create an excessive amount of commits, where humans used to make 10 commits, vibe code agents now often produce 100 or more. This can be solved with our AI Rebase feature.

Select the commits you want to auto rebase, right press with the mouse, and press Auto Rebase. As with the Auto Split feature, we run a recursive algorithm then use AI to decide what logical groups fit best together. Then we reuse the essential commit descriptions and carry them over if relevant. We also use our commit enricher algorithm to further improve descriptions in regards to what each commit now contains. Turn 100 commits into 14 commits that reads more cleanly. The Auto Rebase feature also reduces the amount of file change occurrence significantly.

AI Rebase: Decluttering the Commit Frenzy On the flip side of splitting, there's squashing. The AI Rebase feature lets you select a range of messy commits, and Omni's AI groups them into streamli

eonist

AI Split

For larger refactors with lots of file changes, we can use the Auto Split feature which uses AI to group changes into logical commits. "100 file changes → 12 commits" etc. We use mostly recursive algorithms + AI to figure out the best possible split for the changes given.

Because local models like Apple Foundation Models (AFM) are constrained to tiny context windows (~4K tokens), the system is built on a highly parallel, adaptive "Divide and Conquer" pipeline

eonist

---

How You Can Code Review a PR With Omni

Omni is a great solution to code review with. And while other solutions exist, Omni has the best holistic view of the project out there. Since Omni can pull from 6 context sources, Omni can find nuances and surface things that LLM reviews would miss.

Omni's PR review comment/feature allows users to leverage the Omni project brain to add nuanced reviews with new human intuition. The review is added to the PR comment under the Git user that is currently logged in.

Since Omni can use local AI, you can add a code review to all projects more with out worrying about cost. This can also be done via CLI when vibe coding: omni --pr --review.

eonist

Why We Think Automagic Actions + Human in the Loop is the Future of Git

While vibe code can automatically commit and make PRs, it has its fair share of downsides:

1. Excessive commits on every little thing. 100 commits when a human would need only 10. Especially when instructed to correct things after the fact.
2. Diff-based commit descriptions missing important holistic project knowledge.
3. AIs have limited context windows to deep dive into the intent of the code that was changed.

eonist

How Git Commit, PR and Release Enrichment Works

CLI Commit

Agent + git diff + 4 other enricher sources.

Fetch: git diff, files, websites, connectors, agent input (text transcript)

Process:

eonist

How to Setup Automatic Changelog with GitHub Actions

There is a lot of cool things you can do with CI/CD, such as:

1. Build complex apps on many OSes
2. Build on simple tag change
3. Run tests and security checks
4. Deploy builds to releases that are consumed by users via update automatically
5. Build and ship apps to Apple TestFlight/App Store

eonist

The 6 Context Sources

1. Git Diff

Implicit

Reasoning about changes is pretty trivial. You pass file diffs to an LLM and it returns a sensical description about the change. However, when you want to sort diffs into multiple grouped commits, things get complicated. Here is how we do it.

2. Connectors

Implicit From GitHub we pull issues that are linked via or reference: commit → branch → PR → issue. We look at the issue text and also comments and linked issues/PRs etc.

eonist

---

Why We Built Omni

Changelogs, commit messages, PR descriptions — they're all broken in the age of AI agents. Omni is a dedicated git sub-agent that fixes this by processing six distinct context sources to produce commit history with human-level intuition.

1. Intent → Why It Exists

The problem, the user, the outcome. This defines why the product matters at all.