Ellipsis
AI code review and bug-fixing agents for GitHub pull requests.
About
Ellipsis is an AI code review platform (YC W24) that reviews every pull request as commits land, posting findings as inline comments. Beyond flagging logical bugs, style violations, and antipatterns, its standout feature is turning reviewer comments into working code — developers can ask for a fix in a PR comment and Ellipsis generates a tested commit. It also runs general-purpose cloud coding agents from YAML-defined triggers across GitHub, Linear, Slack, and Sentry.
Pricing is usage-based (model tokens plus compute plus a platform fee) rather than per-seat, which is unusual in this category. New organizations get $100 in one-time credit. The honest caveat: complex architectural changes still need human implementation, and usage-based billing can surprise teams with high review volume. It is GitHub-centric, so GitLab and Bitbucket teams should look elsewhere.
What it can do
- Incremental AI code review on every PR commit
- Turn reviewer comments into tested fix commits
- Custom cloud agents triggered by GitHub, Linear, Slack, or Sentry events
- Natural-language style guide enforcement
- Weekly changelog and release-notes generation
Setup
Install the Ellipsis GitHub App on your repositories from ellipsis.dev. Enable code review with one organization-wide toggle, or define custom agents with YAML pipeline files on your default branch.
Prerequisites
A GitHub organization or personal account. Billing configured once the free credit runs out.
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 44 (tools-a); Trending score 78/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.