GitHub Copilot Chat for Pull Requests Is Now Generally Available
This week, GitHub moved Copilot Chat's richer diff and pull request experience on github.com to general availability. Previously in public preview, the feature is now accessible to anyone holding a Copilot license. The update brings side-by-side chat alongside diffs, faster context-aware answers, and the ability to leave inline comments and edits without toggling away from the pull request view.
It is a meaningful step. The friction of switching between a review diff and a separate chat window is real, and eliminating it reduces the cognitive overhead of getting AI input during review. For teams doing the bulk of their work in a single repository, the experience is measurably tighter.
Why This Matters — and Where It Stops
The problem is that most engineering teams don't work in a single repository. They ship across many — microservices, shared client libraries, infrastructure definitions, platform tooling. A backend change might touch three repos simultaneously. A security patch might need coordinated PRs in five.
In those environments, the quality of feedback on any individual PR is only part of the challenge. The harder part is maintaining awareness of what is open across the organisation at all. A reviewer who doesn't know that a related auth-service PR is sitting in draft, or that a dependent API change was merged two days ago, is reviewing in the dark — regardless of how sharp the in-PR AI assistance is.
GitHub's own research captured this shift well: Gartner notes that by 2028, agentic coding workflows could improve team productivity by 30–50%, but those gains require intelligence across the full SDLC — not just the generation layer, but the review and governance layers where work actually stalls.
The Cross-Repo Review Problem
When a change spans multiple services, the review context lives outside the pull request. It lives in the other PRs that are also in flight — the ones that share a contract, a migration, or a deployment dependency.
This is where current tooling leaves teams on their own. Most review workflows are repo-local by design. Even with AI assistance surfaced inside a PR, there is no native mechanism to surface the cross-repo picture: which related PRs are open, which are approved, which are sitting stale waiting for a review that nobody has gotten to.
The result is familiar to any engineer on a distributed team: tab-hopping across repositories, Slack threads to ask "has anyone merged the config change yet", and PRs that drift stale not because the work is wrong, but because the reviewer lacked context on what else was moving.
This is not a failure of AI tooling. It is a structural gap in how most teams visualise their work in progress.
What Engineering Leaders Should Watch
As AI code review matures — and GitHub's GA of richer Copilot Chat context is a clear signal of the direction — the value floor on per-PR intelligence is rising. Every team will have access to better in-PR feedback. The differentiator will be the teams that can coordinate review across repositories, not just improve it within one.
A few things are worth tracking:
- PR age distribution across repos. Stale PRs are the clearest signal of a review bottleneck. If you cannot see this metric across all repositories in one place, you are guessing about where the friction is.
- Cross-service change correlation. When multiple PRs across services need to land together, the coordination cost is invisible until something breaks. Visibility into simultaneous open PRs across repos surfaces this before it becomes a problem.
- Review load imbalance. AI assistance helps individual reviewers move faster, but it does not redistribute load. A unified view of who is blocked, waiting, or overloaded matters more as teams grow.
The Closing Take
GitHub's move to GA on richer Copilot Chat context in pull requests is a clear sign that AI-assisted review is becoming table stakes. The per-PR experience will keep improving. That is good.
But the teams that gain the most from better per-PR AI are the ones who already have a clear picture of what is in flight across their entire organisation. Without that, sharper in-PR tooling mostly makes individual reviews faster — it does not solve the coordination problem of multi-repo engineering.
For teams working across GitHub and GitLab repositories who want that unified view today, Code Board aggregates every open PR across every repo into a single Kanban board, with AI review built in — so the visibility and the intelligence are in the same place.