GitHub Stacked PRs Are Live: What the Public Preview Actually Changes

GitHub Stacked PRs Are Now Live — Here's What Actually Changed

GitHub stacked PRs moved into public preview on July 30, 2026, and the developer world noticed immediately — the announcement racked up 751 upvotes on Hacker News within hours. For a platform that has historically resisted native support for dependent branch workflows, this is a structural shift, not a cosmetic feature drop. If you've ever split a massive feature into smaller reviewable chunks and then lost a day rebasing five branches after one comment, this changelog entry is for you.

Stacked pull requests let developers break large changes into a sequence of small, dependent PRs — each building on the last — while keeping every layer independently reviewable and mergeable. Until now, doing this on GitHub meant duct-taping the workflow together with tools like Graphite, git-branchless, or manual rebase choreography. GitHub's public preview finally brings this natively into the platform developers already live in.

Why GitHub Stacked PRs Matter Right Now

The core problem stacked PRs solve is review latency. Large PRs sit in review queues longer, get shallower feedback, and produce merge conflicts that compound the longer they wait. Google's internal tooling (Critique) and Meta's Phabricator-based stacks have proven for over a decade that smaller, sequential diffs get reviewed faster and with higher scrutiny — often cutting review cycle time by 30-50% based on internal engineering productivity studies cited by teams like Sourcegraph and Graphite.

GitHub's move essentially validates what a whole ecosystem of third-party tools has been selling as a premium workflow layer. Graphite, founded in 2020 and backed by a $32M Series B round, built its entire product around this exact gap. Native support inside GitHub doesn't kill that ecosystem overnight, but it does put pressure on any tool whose only value proposition was "GitHub doesn't do this."

developer viewing stacked pull request chain.

What's Actually in the Public Preview

Based on the changelog details, the initial public preview focuses on three mechanics: automatic tracking of PR dependencies, a visual stack indicator in the PR sidebar, and smarter rebase handling when an earlier PR in the stack merges. Instead of manually updating base branches across five open PRs, GitHub now propagates the change down the stack automatically. This alone removes one of the most error-prone manual steps in the stacked workflow.

The feature also introduces stack-aware merge queues, meaning CI checks and merge order respect the dependency chain rather than treating each PR as an isolated unit. That's a meaningful upgrade for teams running strict trunk-based development with mandatory status checks, since it prevents a merged parent PR from silently breaking an unmerged child PR still waiting in the queue.

Early preview access appears limited — as is standard for GitHub's public preview rollouts — with broader availability expected to follow the same phased pattern used for Copilot Workspace and merge queue GA in prior release cycles. Enterprise and Team plan customers typically get first access in these rollouts.

The Competitive Pressure on Graphite and Sapling

This is the part of the story most coverage will underplay. Graphite has spent five years building a business on the premise that GitHub's native tooling wasn't sufficient for high-velocity engineering orgs. Meta's Sapling and Graphite's CLI both essentially prove the demand exists — but a native GitHub implementation changes the buying calculus for any team evaluating a third-party subscription purely for stacking mechanics.

That doesn't mean Graphite is obsolete. Its merge queue analytics, AI-assisted review summaries, and cross-repo stack visualization are still ahead of what GitHub has shipped in this initial preview. But the moat just got thinner, and any team currently paying for stacking-as-a-feature should be reassessing whether they still need the external tool or just the workflow.

"Native support from the platform of record always resets the baseline expectation — third-party tools now have to justify their premium beyond the feature GitHub just made free."

How This Changes Day-to-Day Git Workflow

For individual contributors, the practical shift is fewer manual git rebase operations and less mental overhead tracking which branch depends on which. For engineering managers, it means smaller PRs become the path of least resistance instead of the path requiring extra tooling investment. That incentive alignment is what actually drives adoption — teams don't adopt best practices because they're best, they adopt them because the tooling makes them the default.

There's also a review-quality angle worth flagging. Smaller, sequential PRs are demonstrably easier for reviewers to reason about, and GitHub's stacked PR public preview lowers the activation energy required to actually work that way. Combined with GitHub Copilot's PR summary features, the review loop on a five-PR stack could realistically compress from days to hours for well-staffed teams.

engineering team reviewing merge queue dashboard.

Where GitHub Stacked PRs Still Fall Short

Public preview status means rough edges are expected, and the Hacker News thread surfaced several concerns worth tracking. Conflict resolution across a deep stack (five-plus PRs) remains manual in edge cases, particularly when an early PR requires substantial rework after later PRs have already been built on top of it. Some commenters also flagged concerns about how well this integrates with existing branch protection rules and required reviewers across an entire stack.

There's also the migration question for teams already invested in Graphite or Sapling-based workflows — moving a live stack mid-flight between tooling systems isn't trivial, and GitHub hasn't published detailed migration tooling yet. Expect that gap to close over the coming preview cycles based on GitHub's historical pattern with merge queue and Copilot Workspace rollouts.

Should Your Team Switch Right Now?

If you're currently duct-taping stacked workflows with shell scripts or a lighter third-party tool, this preview is worth testing this quarter. If you're a heavy Graphite or Sapling shop with deep integration into your CI/CD and analytics, the calculus is different — evaluate feature parity before ripping out an established workflow. The smart move is running both in parallel on a low-stakes repo before committing org-wide.

side-by-side workflow comparison diagram.

Steps to Get Started With GitHub Stacked PRs

1. Check your organization's plan tier and confirm public preview access is enabled in GitHub's feature settings.
2. Pick one active feature branch with 3+ logical commits and split it into a stack of dependent PRs to test the new tracking.
3. Enable stack-aware merge queue settings and verify your existing branch protection rules still apply correctly across the full stack.
4. Compare your team's average PR review time before and after adopting the stacked workflow over a two-week sprint.
5. Audit any existing Graphite, Sapling, or custom scripts to identify overlap before deciding whether to consolidate tooling.

Frequently Asked Questions

What are stacked PRs on GitHub?

Stacked PRs are a workflow where a large change is split into multiple smaller, sequential pull requests, each depending on the one before it. GitHub's native public preview now tracks these dependencies automatically and handles rebasing and merge order without manual branch management.

Is GitHub's stacked PR feature available to everyone?

As of the July 30, 2026 announcement, it's in public preview, meaning access is rolling out gradually and may be limited to certain plan tiers initially. Broader general availability typically follows GitHub's standard phased rollout pattern seen with prior features like merge queues.

Does this replace tools like Graphite or Sapling?

Not entirely — native support covers core stacking mechanics, but tools like Graphite still offer deeper analytics, AI review summaries, and cross-repo visualization. Teams with heavy investment in those platforms should evaluate feature parity before migrating away.