Ellipsis
AI code review and bug-fix agents that live in your pull requests — YC W24.
About
Ellipsis is an AI developer tool that reviews code and fixes bugs directly on pull requests. When a PR opens, its agents scan for logical errors, security issues, anti-patterns, and style-guide violations — and unlike most review bots, it can implement the fix itself: tag @ellipsis-dev in a comment and it generates, executes, and tests the code, opening the change for your approval.
The product has evolved into a broader governed agent platform: agents run in cloud sandboxes with incremental PR review, triggers, budgets, permissions, and searchable session records, plus a customer-hosted option that keeps credentials, code, and model calls inside your own AWS VPC. The company is YC W24-backed and claims teams merge PRs ~13% faster with it.
The honest caveat: this is closed SaaS, so the speed and quality claims are provider-reported (thinner evidence), and pricing has shifted over time — from per-seat to usage-based billing for tokens and compute with a platform fee. Great for teams that want review automation without self-hosting; less appealing if you need full control or predictable flat pricing.
What it can do
- Automated PR code review: logical bugs, security issues, anti-patterns, style violations
- Agentic bug fixes: generates, executes, and tests code from PR comments
- Natural-language style guides enforced automatically across reviews
- Cloud sandboxes with budgets, permissions, triggers, and searchable session records
- Customer-hosted option running inside your own AWS VPC
- GitHub + GitLab + Slack integrations; 20+ languages supported
Setup
- Install the Ellipsis GitHub app on your repositories (two-click install; free trial available).
- Configure review triggers, permissions, and budgets in the Ellipsis dashboard.
- Open a PR — Ellipsis reviews it automatically and comments with findings.
- Tag
@ellipsis-devin any PR comment to ask questions or request fixes; approve changes before merge.
Prerequisites
A GitHub organization or account; billing configured after the trial/credits run out (new orgs get usage credit).
Use cases
Works with
Trust score — 7/10
Scored 7/10 (verified tier, metrics-v1). Evidence: Adoption 6: closed SaaS, no repo — estimated from YC W24 backing, 9,300+ developers / 29,000 codebases per third-party directories; thinner evidence; Maintenance 8: active company, docs current; SaaS maintenance estimated, not repo-derived; Manual 7: provider speed/quality claims (13% faster merges) treated as directional; pricing model has shifted over time — flagged. License: Closed SaaS (no public repo).
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 6.0/10
- Maintenance · recency, releases, issue load
- 8.0/10
- Openness · license + docs
- 5.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · Adoption 6: closed SaaS, no repo — estimated from YC W24 backing, 9,300+ developers / 29,000 codebases per third-party directories; thinner evidence
- · Maintenance 8: active company, docs current; SaaS maintenance estimated, not repo-derived
- · Manual 7: provider speed/quality claims (13% faster merges) treated as directional; pricing model has shifted over time — flagged
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.