Every source behind the read

38 sources in Engineering leaders and teams, with product documentation Showing all 66 sources

1. Positive

   Dotta, cofounder of Paperclip

   Ep. #45, When Agents Outrun Human Review with Dotta of PaperclipPodcast episode transcript, Open Source Ready (Heavybit)

   Paperclip's cofounder calls Greptile critical to the project's infrastructure for screening about 5,000 community pull requests. Attackers wait for Greptile and the other checks to pass and then push a commit with the exploit, so the team reruns every check on each new commit at real inference cost.

   "…we use Greptile, which I love. Not paid at all. It's critical to our infrastructure. We're leaning hard on Greptile."

   "Greptile and all our checks will run it and it'll pass. And as soon as it passes, they submit another commit that introduces the hack essentially."

   Published Oct 22, 2026Paperclip, an open-source agent platform that has received about 5,000 community pull requests

   Counts toward

   • Overall signal
   • Workflow Fit: Positive
   • GitHub Workflow Friction: Discussed without concern
   • After a trial or a price change, do teams keep Greptile?: Kept it

   Published on

   Talks & Podcasts

   Who

   Technical leadership

   Evidence

   Reported

   Applies to

   Current

2. Positive

   Mike Cao, runs the start-up Umami

   Post by @caozillaX post

   The head of Umami says the team uses Greptile every day on Umami and calls it an awesome product.

   "Greptile is such an awesome product. We use it every day in Umami."

   Published Oct 20, 2026Umami, open source web analytics company

   Counts toward

   • Overall signal
   • Workflow Fit: Positive
   • After a trial or a price change, do teams keep Greptile?: Kept it

   Published on

   Developer Communities

   Who

   Technical leadership

   Evidence

   Reported

   Applies to

   Current

3. Positive

   Nick Nisi, engineering manager on the DX / AI team at WorkOS

   Riker: An Agent in My AgentPersonal site post, nicknisi.com

   A WorkOS engineering manager no longer reads full diffs before merging agent PRs: he reviews only the parts flagged as riskier, and on the latest PR he checked that it got good Greptile scores and then checked the proof screenshots. Since September 25, his agent manager has run 141 jobs, and 94 of the resulting PRs were merged.

   "On the last one, I simply checked that it got good scores from Greptile and then I checked the proof."

   "Before I merge something, I rely on the update + the review tab to look at any parts of the code that are classified as riskier, but definitely not the full PR."

   Published Oct 12, 2026WorkOS DX and AI team; an internal agent manager hands work to coding agents and returns PRs with proof

   Counts toward

   • Overall signal
   • Summaries & Confidence Scores: Positive
   • A first pass before human review
   • After a trial or a price change, do teams keep Greptile?: Kept it

   Published on

   Independent Reviews

   Who

   Technical leadership

   Evidence

   Reported

   Applies to

   Current

   Observed on

   Greptile scores on PRs opened by Riker, his agent manager, since September 2026

4. Positive

   Torsten Dittmann, product architect at Appwrite

   Don't approve while a bug finding is open, and re-check approvals after new commitsGitHub issue

   On an Appwrite pull request, Greptile found two serious bugs that the author confirmed and fixed, after the author's own reviewer, Hansi, had approved the change and later missed one of them.

   Published Oct 6, 2026self-described: Appwrite's appwrite/appwrite pull requests, where the author compares a self-built reviewer, Hansi, with Tenki and Greptile

   Counts toward

   • Overall signal
   • Bug Detection: Positive
   • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers Greptile

   Published on

   Developer Communities

   Who

   Professional team

   Evidence

   Reported

   Applies to

   Current

5. Positive

   Contributor to pomerium/agentops

   Platform stack: Greptile findings left after five review roundsGitHub issue

   Pomerium's agentops stack went through five Greptile review rounds, each fixed test-first. The issue lists the concurrency and error-handling findings still open, such as queued startup errors that are dropped and retries that outlast a token.

   Published Oct 6, 2026self-described: Pomerium's agentops repository, a stack of 12 pull requests reviewed by Greptile

   Counts toward

   • Overall signal
   • Bug Detection: Positive
   • A first pass before human review
   • After a trial or a price change, do teams keep Greptile?: Kept it

   Published on

   Developer Communities

   Who

   Professional team

   Evidence

   Reported

   Applies to

   Current

   Observed on

   Greptile GitHub app with 5/5 confidence scores

6. Mixed

   Georges-Antoine Assi, engineer at PostHog

   chore: retire coderabbit reviews, skip greptile on skip-agent-reviewGitHub pull request on PostHog/posthog

   PostHog switched CodeRabbit off in favor of its own reviewer but chose to keep Greptile's automatic reviews, pointing Greptile's disabledLabels setting at the shared skip-agent-review label because Greptile ignored it before. The team deleted a nested desktop config since a nested .greptile folder inherits the root config.

   "The reviewer then chose to leave Greptile's automatic reviews on, to point its skip label at skip-agent-review, and to drop the redundant desktop config."

   Published Oct 6, 2026self-described: PostHog's main repository, where the company retired CodeRabbit for its in-house reviewer

   Counts toward

   • Overall signal
   • GitHub Workflow Friction: Discussed without concern
   • After a trial or a price change, do teams keep Greptile?: Kept it

   Published on

   Developer Communities

   Who

   Organization-wide

   Evidence

   Reported

   Applies to

   Current

   Observed on

   Root .greptile config with disabledLabels; nested .greptile folder in products/desktop

7. Mixed

   Owner of Youtab B.V.'s Youtab-Agent-Runtime repository

   Owner governance: replace Greptile with Independent Agent ReviewGitHub issue

   Retired Greptile as a merge and release gate in favor of a read-only review by an agent that did not write the change. Codex is the final technical reviewer, and past Greptile reviews stay on record.

   Published Sep 10, 2026self-described: Youtab B.V., an agent runtime repository with required merge gates

   Counts toward

   • Overall signal
   • After a trial or a price change, do teams keep Greptile?: Dropped it
   • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers another reviewer

   Published on

   Developer Communities

   Who

   Professional team

   Evidence

   Reported

   Applies to

   Current

8. Positive

   Luca Marturana, Chief Architect at Spark

   [Technical] Code reviews at Spark: What we changed after AICompany blog post, SparkNotes (Spark)

   Spark's Chief Architect reports that the team uses Greptile to catch bugs in review and that it has caught both large and small ones, such as an unhandled business-logic edge case and a missing data migration step. Style rules run through an in-house Claude Code agent, and humans review design documents, data models and customer fit.

   "We leverage Greptile for the former; in our experience, it has been very good at catching bugs big and small, such as an edge case in the business logic we forgot to handle, or a data migration step we forgot to add."

   Published Sep 9, 2026self-described: Spark, early-stage energy software company; agents write most code, Claude Code style agent beside Greptile

   Counts toward

   • Overall signal
   • Bug Detection: Positive
   • A first pass before human review
   • After a trial or a price change, do teams keep Greptile?: Kept it

   Published on

   Independent Reviews

   Who

   Technical leadership

   Evidence

   Reported

   Applies to

   Current

   Observed on

   Greptile for bug detection; an in-house Claude Code agent for house style rules

9. Positive

   Sherwood, founder of Sazabi (X profile)

   Post by @shcallawayX post

   After trying Bugbot, Graphite, CodeRabbit, Codex and others, calls Greptile their favorite reviewer and says it rarely lets a bug or regression through.

   Published Sep 8, 2026self-described: Sazabi engineering team

   Counts toward

   • Overall signal
   • Bug Detection: Positive
   • Summaries & Confidence Scores: Positive
   • After a trial or a price change, do teams keep Greptile?: Kept it
   • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers Greptile

   Published on

   Developer Communities

   Who

   Technical leadership

   Evidence

   Reported

   Applies to

   Current

10. Mixed

    Maintainer of deftai/directive

    fix(ci): Dependabot PRs are unmergeable without admin because Greptile posts no required checkGitHub issue

    Their organization requires a Greptile Review check before merge, but Greptile posts no check run for authors it excludes, so every Dependabot pull request needs an admin to bypass the rule.

    Published Aug 24, 2026self-described: deftai organization with a ruleset that requires a Greptile Review check on the default branch

    Counts toward

    • Overall signal
    • Workflow Fit: Mixed or negative
    • GitHub Workflow Friction: Raised as a concern
    • A first pass before human review

    Published on

    Developer Communities

    Who

    Professional team

    Evidence

    Reported

    Applies to

    Current

    Observed on

    Greptile GitHub app as a required status check, Dependabot in the excluded-authors list

11. Positive

    Andrew Churchill, CTO of Weave

    3 things top 1% teams do differentlyNewsletter guest post, Manager.dev

    Weave's CTO runs Greptile with Cursor Bugbot and Cubic on every pull request because each catches bugs the others miss, and treats a PR with no meaningful bot comment as free of obvious bugs. Human review now goes mostly to plans and to sensitive areas such as infra.

    "We use three different ones: Cursor Bugbot, Cubic and Greptile. Each of them catches some stuff that the other ones don't, and as long as that is true, it is worth it to us to have 3 of them."

    "If there is no meaningful comment from any AI bot, we can be pretty confident there are no obvious bugs in it."

    Published Jul 28, 2026Weave engineering team; three AI reviewers on every PR, about 5% of PRs get a human review

    Counts toward

    • Overall signal
    • Bug Detection: Positive
    • A first pass before human review
    • After a trial or a price change, do teams keep Greptile?: Kept it

    Published on

    Independent Reviews

    Who

    Technical leadership

    Evidence

    Reported

    Applies to

    Current

    Observed on

    Greptile alongside Cursor Bugbot and Cubic

12. Mixed

    Maintainer of Brite Nites' Claude Code plugins

    greptile-gate: greptile-await.sh false TIMED_OUT, Greptile edits its summary comment in place (createdAt stays pre-trigger)GitHub issue

    Their team gates merges on Greptile's score. Because Greptile edits its summary comment in place on a re-review, the team's wait script timed out even after the score had moved from 4/5 to 5/5.

    Published Jul 8, 2026self-described: Brite Nites, a merge gate that waits for Greptile's score on the brite-data-platform repository

    Counts toward

    • Overall signal
    • Workflow Fit: Mixed or negative
    • GitHub Workflow Friction: Raised as a concern
    • A first pass before human review

    Published on

    Developer Communities

    Who

    Professional team

    Evidence

    Reported

    Applies to

    Current

    Observed on

    Greptile GitHub app with confidence scores used as a merge gate

13. Mixed

    ElFitz, senior software engineer at Purchasely (HN profile)

    Comment on a Hacker News thread about human effort in pull requestsHacker News comment

    Runs Greptile alongside Codex, Gemini, Copilot and CodeRabbit on every pull request. The reviewers rarely overlap yet usually miss the same things, so human review can focus elsewhere.

    Published Jun 12, 2026self-described: Purchasely, multi-model AI review on every pull request

    Counts toward

    • Overall signal
    • Missed Issues: Raised as a concern
    • A first pass before human review

    Published on

    Developer Communities

    Who

    Professional team

    Evidence

    Reported

    Applies to

    Current

14. Negative

    sebra, Hacker News commenter

    Comment on: There is an AI code review bubbleHacker News comment

    Ran Greptile on three pull requests and stopped: comments missed context handled elsewhere, flagged Python 3.14 as nonexistent, and the confidence score misled reviewers; CodeRabbit had worked better.

    Published Jan 27, 2026self-described: Employer team, Python codebase, previously on CodeRabbit

    Counts toward

    • Overall signal
    • Codebase Context: Mixed or negative
    • Summaries & Confidence Scores: Mixed or negative
    • Comment Noise: Raised as a concern
    • After a trial or a price change, do teams keep Greptile?: Dropped it
    • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers another reviewer

    Published on

    Developer Communities

    Who

    Professional team

    Evidence

    Reported

    Applies to

    Current

15. Mixed

    disillusionist, Hacker News commenter

    Comment on: There is an AI code review bubbleHacker News comment

    After a several-week trial against Cursor, their company declined to buy Greptile. Developers were split, and it overwrote PR descriptions and argued with itself in comments.

    Published Jan 26, 2026self-described: Employer engineering team, evaluating against Cursor

    Counts toward

    • Overall signal
    • Bug Detection: Positive
    • Summaries & Confidence Scores: Mixed or negative
    • Comment Noise: Raised as a concern
    • After a trial or a price change, do teams keep Greptile?: Dropped it
    • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers another reviewer

    Published on

    Developer Communities

    Who

    Professional team

    Evidence

    Reported

    Applies to

    Current

16. Negative

    kaishin, Hacker News commenter

    Comment on: There is an AI code review bubbleHacker News comment

    Their workplace used Greptile and replaced it with Claude, though Claude still falls short of an experienced programmer with domain knowledge.

    Published Jan 26, 2026self-described: Employer team

    Counts toward

    • Overall signal
    • A first pass before human review
    • After a trial or a price change, do teams keep Greptile?: Dropped it
    • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers another reviewer

    Published on

    Developer Communities

    Who

    Professional team

    Evidence

    Reported

    Applies to

    Current

17. Positive

    James Reggio, CTO at Brex

    Full Story of Brex’s AI Hail MaryPodcast episode show notes and transcript, Latent Space

    Brex aligned on Greptile as its one agentic code review tool on top of traditional linters. Its CTO singles out the signal-to-noise ratio of its pull request comments and says he never regrets reading all 65 comments it leaves on his diffs.

    "According to James, what is most impressive about Greptile is their signal to noise ratio of comments left on a PR, “I never regret going through all 65 comments [Greptile] leaves on my diffs because it catches so many things.”…"

    Published Jan 17, 2026Brex, about 300 engineers; traditional linters plus Greptile for agentic code review

    Counts toward

    • Overall signal
    • Bug Detection: Positive
    • Comment Noise: Discussed without concern
    • After a trial or a price change, do teams keep Greptile?: Kept it

    Published on

    Talks & Podcasts

    Who

    Organization-wide

    Evidence

    Reported

    Applies to

    Current

18. Negative

    CSMastermind, Hacker News commenter

    Comment on: Codemaps: Understand Code, Before You Vibe ItHacker News comment

    Ranks Codex's pull request review bot well ahead of the alternatives their team tried, Greptile among them.

    Published Nov 5, 2025self-described: Team that compared several review bots

    Counts toward

    • Overall signal
    • After a trial or a price change, do teams keep Greptile?: Dropped it
    • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers another reviewer

    Published on

    Developer Communities

    Who

    Professional team

    Evidence

    Reported

    Applies to

    Still applicable

19. Positive

    Ovais Tariq, co-founder and CEO at Tigris Data (HN profile)

    Comment on: AI tooling must be disclosed for contributionsHacker News comment

    Argues that growing AI-generated code should be handled through review tooling, and reports good results from Greptile, Cursor's Bugbot and Claude Code.

    Published Aug 21, 2025self-described: Tigris Data

    Counts toward

    • Overall signal
    • After a trial or a price change, do teams keep Greptile?: Kept it

    Published on

    Developer Communities

    Who

    Technical leadership

    Evidence

    Reported

    Applies to

    Still applicable

20. Positive

    James Hartig, software engineer at Admiral (HN profile)

    Comment on: Why Greptile just does code reviews and doesn't also generate codeHacker News comment

    Their employer uses Greptile. It is sometimes overzealous but usually catches real issues, so authors fix them before a human reviews the pull request.

    Published Aug 4, 2025self-described: Admiral, engineering team

    Counts toward

    • Overall signal
    • Bug Detection: Positive
    • Comment Noise: Discussed without concern
    • A first pass before human review
    • After a trial or a price change, do teams keep Greptile?: Kept it

    Published on

    Developer Communities

    Who

    Professional team

    Evidence

    Reported

    Applies to

    Still applicable

21. Positive

    Repository owner

    feat(dev-loop): add Greptile and Graphite (free plans) to the review ladder after CodeRabbitGitHub issue

    Added Greptile's free Starter plan as a fallback reviewer after rate limits left about 15 pull requests unreviewed for four hours. Once installed, it reviewed new pull requests on open with inline threads and severity badges.

    Published Oct 11, 2026self-described: One-maintainer setup whose agent loop falls back across several hosted and local reviewers

    Counts toward

    • Overall signal
    • Workflow Fit: Positive
    • Pricing & Billing: Discussed without concern
    • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers Greptile

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

    Observed on

    Greptile Starter plan: free, 50 credits a month, one active developer

22. Positive

    Piyush T., reviewer on G2, self-described

    Deep, Codebase-Wide Reviews That Catch More IssuesG2 review, syndicated on AWS Marketplace

    Rates it 4.5 of 5 for understanding the whole codebase beyond the changed lines. Deeper reviews add latency and usage cost on large pull requests.

    Published Oct 6, 2026

    Counts toward

    • Overall signal
    • Codebase Context: Positive
    • Pricing & Billing: Discussed without concern

    Published on

    Independent Reviews

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

23. Positive

    Rasmita J., reviewer on G2

    Deep Codebase Understanding with Repo Graph, Bug & Security DetectionG2 review, syndicated on AWS Marketplace

    Rates it 4.5 of 5 for predicting a change's effect on the rest of the repository and catching bugs and security issues. Comments improve as it learns the team's patterns. But reviews are slow, and the new pricing costs more than competitors as pull request volume grows.

    Published Sep 25, 2026self-described: Own team's repositories

    Counts toward

    • Overall signal
    • Bug Detection: Positive
    • Codebase Context: Positive
    • Comment Noise: Discussed without concern
    • Pricing & Billing: Raised as a concern

    Published on

    Independent Reviews

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

    Observed on

    Per-developer pricing after March 2026

24. Negative

    Grimblaz, maintainer of tradecraft

    A connected reviewer we own replaces GreptileGitHub issue

    Removed Greptile after its trial ran out to avoid paying US$30 a month plus US$1 per review. Designed a Claude-based reviewer.

    Published Sep 24, 2026self-described: Own repositories reviewed by Greptile and Codex

    Counts toward

    • Overall signal
    • Pricing & Billing: Raised as a concern
    • After a trial or a price change, do teams keep Greptile?: Dropped it
    • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers another reviewer

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

25. Mixed

    Viktor Farcic, maintainer of dot-agent-deck

    Compare Qodo and Greptile reviews and pick oneGitHub issue

    Runs Qodo beside Greptile and plans a head-to-head on valid findings and noise. Greptile's monthly quota forced Farcic to stop reviews on every push.

    Published Sep 23, 2026self-described: Open source repository reviewed by Greptile and Qodo

    Counts toward

    • Overall signal
    • Comment Noise: Discussed without concern
    • Pricing & Billing: Raised as a concern

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

    Observed on

    greptile.json with triggerOnUpdates false

26. Mixed

    Tuli D., reviewer on G2

    Seamless GitHub Integration with Hands-OffG2 review, syndicated on AWS Marketplace

    Setup and automatic reviews on every pull request were straightforward. Three test pull requests, one with a deliberate division-by-zero bug, all scored 4 of 5 with low severity, which questions how granular the score is.

    Published Sep 19, 2026self-described: Test pull requests in Python and JavaScript

    Counts toward

    • Overall signal
    • Summaries & Confidence Scores: Mixed or negative
    • Workflow Fit: Positive
    • A first pass before human review

    Published on

    Independent Reviews

    Who

    Individual practitioner

    Evidence

    Measured

    Applies to

    Current

27. Mixed

    Izzy H., reviewer on G2

    Speeds Up Understanding Large Codebases, Though Questions Need Careful PhrasingG2 review, syndicated on AWS Marketplace

    Rates it 3.5 of 5: asking questions about a large unfamiliar codebase gives a quick starting point, but answer quality depends on how the question is phrased.

    Published Sep 15, 2026self-described: Large unfamiliar codebase

    Counts toward

    • Overall signal
    • Codebase Context: Positive

    Published on

    Independent Reviews

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

28. Mixed

    acharlop, whocards platform maintainer

    Measure Greptile against the code-review skill: per-PR diff of findingsGitHub issue

    Reactivated Greptile on the free tier after its earlier output had real P1 catches but mostly hygiene comments, and set up a per-PR comparison to decide whether to keep it.

    Published Sep 7, 2026self-described: whocards platform repository, free tier

    Counts toward

    • Overall signal
    • Bug Detection: Positive
    • Comment Noise: Raised as a concern

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

    Observed on

    Starter free tier, 50 reviews a month

29. Positive

    Senior Rails engineer, Levelbrook Consulting (HN profile)

    Comment on: Ask HN: Who wants to be hired? (September 2026)Hacker News comment

    Credits an agentic loop of Claude Code, Greptile, GitHub Actions, and Playwright with tripling their year-over-year pull request output on a large production codebase.

    Published Sep 2, 2026self-described: Contract work on a large production codebase

    Counts toward

    • Overall signal
    • Workflow Fit: Positive
    • After a trial or a price change, do teams keep Greptile?: Kept it

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

30. Positive

    Kirpalsinh R., reviewer on G2

    Context-aware AI code reviews that actually understand the whole repositoryG2 review, syndicated on AWS Marketplace

    Rates it 5 of 5 for judging cross-file impact from a graph of the repository and catching logic and security issues early. But it is noisy on large pull requests until the team trains it with reactions, and a single pass can miss edge cases.

    Published Sep 2, 2026self-described: Team use

    Counts toward

    • Overall signal
    • Bug Detection: Positive
    • Codebase Context: Positive
    • Comment Noise: Raised as a concern
    • Missed Issues: Raised as a concern
    • A first pass before human review

    Published on

    Independent Reviews

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

31. Positive

    Prateek M., reviewer on G2

    Detailed PR Feedback, Fast Analysis, and Smooth GitHub IntegrationG2 review, syndicated on AWS Marketplace

    Rates it 5 of 5 after long use: detailed pull request comments with a fix-with-Claude button and a first pass that frees senior reviewers. Pricing trails competitors, and it misses the overall picture of a pull request.

    Published Aug 27, 2026self-described: Team use over a long period

    Counts toward

    • Overall signal
    • Workflow Fit: Positive
    • Pricing & Billing: Raised as a concern
    • A first pass before human review
    • After a trial or a price change, do teams keep Greptile?: Kept it

    Published on

    Independent Reviews

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

32. Positive

    m463, Hacker News commenter, self-described

    Comment on: Reviews have become expensive, rewrites have become cheapHacker News comment

    Found Greptile's review decent, closer to a static analysis tool without the time-wasting false positives, while design questions still come from human developers.

    Published Jun 16, 2026

    Counts toward

    • Overall signal
    • Comment Noise: Discussed without concern
    • A first pass before human review

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

33. Positive

    3dsnano, Hacker News commenter

    Comment on: Ask HN: Is Anybody Using Codex?Hacker News comment

    Describes a loop where Claude implements and Codex reviews, then Greptile reviews the commit and the /greploop command closes out Greptile's remaining comments.

    Published Apr 30, 2026self-described: Solo workflow with Claude Code and Codex

    Counts toward

    • Overall signal
    • Workflow Fit: Positive
    • After a trial or a price change, do teams keep Greptile?: Kept it

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

34. Negative

    epistasis, Hacker News commenter, self-described

    Comment on: Small models also found the vulnerabilities that Mythos foundHacker News comment

    From experience with Greptile, expects its false positive rate to be too high for whole-codebase scanning to be useful.

    Published Apr 11, 2026

    Counts toward

    • Overall signal
    • Comment Noise: Raised as a concern

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

35. Negative

    kusokurae, Hacker News commenter, self-described

    Comment on: Slop is not necessarily the futureHacker News comment

    Says Greptile floods pull requests with verbose comments and often makes recommendations that would be harmful to apply.

    Published Apr 1, 2026

    Counts toward

    • Overall signal
    • Comment Noise: Raised as a concern

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

36. Mixed

    raflueder, Hacker News commenter

    Comment on: Code Review for Claude CodeHacker News comment

    After experimenting with Greptile on a solo project, moved to a self-built Claude-based review workflow that averaged US$0.04 a review over February.

    Published Mar 9, 2026self-described: Solo project

    Counts toward

    • Overall signal
    • After a trial or a price change, do teams keep Greptile?: Dropped it
    • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers another reviewer

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

    Observed on

    Compared with a homegrown review workflow on Opus 4.6

37. Positive

    Nik Cubrilovic, developer (HN profile)

    Comment on: Orchestrate teams of Claude Code sessionsHacker News comment

    Codes with Opus, reviews each task with Codex through a review skill, and also runs Greptile on every pull request.

    Published Feb 5, 2026self-described: Own projects, Opus for coding and Codex for review

    Counts toward

    • Overall signal
    • After a trial or a price change, do teams keep Greptile?: Kept it

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

38. Negative

    Kye, builds agents at Falconer (X profile)

    Post by @kyeburchardX post

    On one pull request, Greptile caught 1 of 3 real bugs against 2 for Codex and 3 for Cursor's Bugbot, which alone caught a bug needing context from another file.

    Published Jan 30, 2026self-described: Single pull request comparison

    Counts toward

    • Overall signal
    • Bug Detection: Mixed or negative
    • Codebase Context: Mixed or negative
    • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers another reviewer

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Measured

    Applies to

    Current

39. Negative

    iblaine, data engineer (HN profile)

    Comment on: There is an AI code review bubbleHacker News comment

    Had a bad experience with Greptile because of excessive noise and nitpick comments, and has used Cursor's Bugbot for a year instead.

    Published Jan 27, 2026self-described: Data engineer

    Counts toward

    • Overall signal
    • Comment Noise: Raised as a concern
    • After a trial or a price change, do teams keep Greptile?: Dropped it
    • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers another reviewer

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

40. Positive

    mohsen1, Hacker News commenter

    Comment on: There is an AI code review bubbleHacker News comment

    Tried Copilot and Cubic for code review and settled on Greptile, and is happy with it.

    Published Jan 26, 2026self-described: Own repositories

    Counts toward

    • Overall signal
    • After a trial or a price change, do teams keep Greptile?: Kept it
    • Against other AI reviewers tried hands-on, does Greptile come out ahead?: Prefers Greptile

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

41. Mixed

    seanmccann, Hacker News commenter

    Comment on: There is an AI code review bubbleHacker News comment

    Reports that Greptile finds fewer issues in their pull requests as Claude Code and Opus improve.

    Published Jan 26, 2026self-described: Own code written with Claude Code

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

    Observed on

    Code written with Claude Code and Opus

42. Positive

    ej88, Hacker News commenter

    Comment on: The creator of Claude Code's Claude setupHacker News comment

    Has Claude Code respond to Greptile reviews automatically before a pull request reaches a human, and still reviews every pull request personally, which is the bottleneck.

    Published Jan 7, 2026self-described: Solo, 3 to 5 parallel Claude Code sessions

    Counts toward

    • Overall signal
    • Workflow Fit: Positive
    • A first pass before human review

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

43. Mixed

    Michael, member of technical staff at context.dev (X profile)

    Post by @michael_chomskyX post

    After weeks running Greptile and CodeRabbit together, found Greptile more focused and its 5/5 scores useful for merging, while both flag non-issues and miss critical problems.

    Published Dec 22, 2025self-described: Own projects on a Bun, oRPC and better-auth stack; free access from both vendors

    Not counted

    Its author received free access from Greptile. Kept in the record; it takes part in no figure.

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Still applicable

    Observed on

    Free access to Greptile and CodeRabbit

44. Positive

    Alex Chantavy, open source maintainer (HN profile)

    Comment on: Potential issues in curl found using AI assisted toolsHacker News comment

    As an open source maintainer, finds that AI reviewers such as cubic and Greptile save time, catch things a human would have missed, and help when contributor quality varies.

    Published Oct 2, 2025self-described: Maintains an open source project

    Counts toward

    • Overall signal
    • Bug Detection: Positive
    • A first pass before human review

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Current

45. Negative

    bwfan123, Hacker News commenter, self-described

    Comment on: AI-generated "workslop" is destroying productivity?Hacker News comment

    Says CodeRabbit and Greptile produce large volumes of review comments that take real effort to read, for an occasional useful finding.

    Published Sep 22, 2025

    Counts toward

    • Overall signal
    • Bug Detection: Mixed or negative
    • Comment Noise: Raised as a concern

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Still applicable

46. Positive

    Emily T. Burak, developer

    Greptile -- AI Pull Request Reviews in 2025Personal blog post

    On a personal Go and HTMX project, Greptile used the pull request description as context and caught a hardcoded port, a prematurely closed database connection, and table name mismatches. Burak sees its review as tactical and finds it lacks a strategic perspective.

    Published May 25, 2025self-described: Solace, a personal Go and HTMX project; Greptile with default settings beside GitHub Copilot and Cursor

    Counts toward

    • Overall signal
    • Bug Detection: Positive
    • Codebase Context: Positive
    • A first pass before human review

    Published on

    Independent Reviews

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Still applicable

    Observed on

    Greptile pull request review with default greptile.json settings

47. Negative

    clintonb, Hacker News commenter

    Comment on: Claude IntegrationsHacker News comment

    Recalls being annoyed by the suggestions Greptile posted as comments on Linear bug tickets, a feature Greptile later dropped to focus on code review.

    Published May 2, 2025self-described: Greptile comments on bug tickets in Linear

    Counts toward

    • Overall signal
    • Comment Noise: Raised as a concern

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Still applicable

    Observed on

    Greptile's Linear integration for bug tickets, since deprecated

48. Positive

    999900000999, Hacker News commenter

    Comment on: Kilo Code: Speedrunning open source coding AIHacker News comment

    Says Greptile works well because it ingests the whole repository, so on a Haxe framework it answers far better than ChatGPT, at the cost of digesting the repository first.

    Published Mar 26, 2025self-described: Feeding Greptile a Haxe framework repository

    Counts toward

    • Overall signal
    • Codebase Context: Positive

    Published on

    Developer Communities

    Who

    Individual practitioner

    Evidence

    Reported

    Applies to

    Still applicable