Code review for your architecture.

Not file diffs. Not text. Striff reviews every pull request as a dependency graph — boundary violations, coupling growth, and package cycles surfaced before merge.

What posts on your pull request — a diagram of what changed, plus ranked findings
apache/fineract PR #6063 Java 2 HIGH · 2 MEDIUM
Striff / architectural-review
Striff structural diff for apache/fineract PR #6063
Scroll to zoom · Drag to pan
Architecture diagram of this PR
HIGH
Package structure violation

DefaultCommandHandlerManager sits in implementation, but REFACTORING.md requires command handlers to live in handler. Doc-grounded — cites the file by name.

HIGH
New directional boundary crossing

New dependency from command to command.implementation — no prior edge in this direction.

MED
Complexity growth: WMC +10

DefaultCommandHandlerManager WMC grew from 4 to 14 — rising cyclomatic complexity in the misplaced class.

MED
Coupling spike: EC +9

DefaultCommandHandlerManager efferent coupling grew from 8 to 17 — harder to change in isolation.

No other tool reads your docs. Striff cites REFACTORING.md by name and flags the exact class that violates it.

The problem

Architecture matters more, not less.

Your team merges more AI-written code every week than anyone can read. Left unchecked, each merge piles on coupling and cycles until the architecture is unmaintainable. Striff catches that structural drift on every pull request.

Why it matters more in the AI era, not less — read the breakdown.

Research-backed — roots at ICCQ 2021

The proof, on live PRs

The riskiest changes look safe in a file diff.

These are real public pull requests. Each one looked routine file-by-file. Striff read the structure underneath — a new package cycle, a god-class forming, a migration quietly reversing — and posted the findings before merge.

  • Live public PRs — open any of them and verify
  • Findings posted on the PR before merge
  • Zero overlap with AI reviewers' comments

Built for engineering leaders

Fast teams. Architecture that holds.

Architectural drift is the tax your team pays later — slower PRs, harder onboarding, rewrites nobody budgeted for. Striff checks every PR against the architecture you've already documented, rolls it into an org-wide health dashboard, and never slows a single review down to do it.

Striff structural diagram for apache/fineract PR #6063, zoomed on DefaultCommandHandlerManager
HIGH
Package structure violation

DefaultCommandHandlerManager sits in implementation, but REFACTORING.md requires command handlers to live in handler. Doc-grounded — cites the file by name.

Last 6 monthsFebruary 2026 -- July 2026

Regressions flaggedA high-severity structural break Striff traced directly to this PR -- a new dependency cycle, a one-way boundary crossing, a stable component's contract shifting, or a sharp complexity jump. Deliberately rare: most PRs show zero.

30

Hotspots flaggedA lower-severity or anomaly-only finding worth a second look -- coupling or churn signals that don't rise to a confirmed structural regression. Usually zero or one per PR.

54

Clean PR rateShare of analyzed pull requests with zero regressions or hotspots flagged, over the last 6 months.

75%

High-risk PR rateShare of analyzed pull requests with at least one regression flagged, the more severe finding type, over the last 6 months.

7%

PRs analyzedTotal pull requests Striff reviewed for architecture across every active repo in this installation.

425

CoverageShare of GitHub PR-check webhook events (opened, updated, reopened) that completed analysis, over the last 6 months. Below 100% may mean PRs were skipped -- check billing status or repo connection.

93%

Flagged repos over timeEvery repo that has cracked the top-flagged list at any point in the last 6 months, tracked month by month. A repo can show a lower or zero count in months it wasn't flagged enough to be in that month's own top list -- more than one repo commonly appears here as the top spot reshuffles across months.

Recently flagged PRsThe 10 most recent pull requests this month with a structural regression or review hotspot -- click through to see exactly what was flagged.

Jul 2026

acme-corp / platform-api

Add token-bucket rate limiter to gateway middleware #1842

Open jordan-lee wants to merge 6 commits into main from feature/rate-limiter
Conversation 12 Commits 6 Checks 4 Files changed 9

All checks have passed

4 successful checks

continuous-integration / build 44s Details
continuous-integration / test (312) 1m 12s Details
lint 18s Details
Striff — architecture

No boundary violations. Rate limiter stays inside gateway/middleware; matches ARCHITECTURE.md.

26s Details

Free Chrome extension

See the architecture, right inside GitHub.

Install it and a new "Striffs" tab shows up next to Files changed on any public PR — no account, no backend to run.

Live architecture diagrams, right on the PR

A new "Striffs" tab appears next to GitHub's Files changed. Click it and get a real, pannable, zoomable dependency diagram of the change — not a static screenshot.

Click a component, drop a comment

Hover any class or package in the diagram, hit +, and post a real PR comment with a rendered sub-diagram attached — no screenshotting, no leaving the page.

AI notes overlaid directly on the graph

Architectural findings sit right on the diagram entities they describe, with inline thumbs up/down so your team can rate findings in place.

Zero setup, instant on any public repo

No backend to stand up, no account required to start. Install and go — repeat visits to the same PR load instantly from cache.

Get started

Start with public repos. Scale to your team.

Striff runs as a GitHub Check. No new tool to learn, no workflow to change. Start free on public repos, add private repos when you're ready.

Free

$0 forever

Public repositories

Browser extension for any public PR. Structural diffs, unusual-edge signals, and AI review notes.

Starter

$29 /mo · 5 repos

Private repos

GitHub App checks on every private PR. All detectors, GNN scoring, and automated review.

Team $59/mo · Scale $149/mo →

Enterprise

Talk to us

Self-hosted & 50+ repos

Private deployment, SSO, and audit workflows for organizations that need full control.

FAQ

Common questions

Does Striff replace CodeRabbit, Copilot, or Claude?

No — Striff complements them. AI reviewers catch code-level bugs, style violations, and logic errors. Striff catches architectural risks: dependency cycles, coupling spikes, boundary violations. Use them together for full coverage.

What languages does Striff support?

Java, TypeScript, Python, and C#. Go support is coming soon. Striff parses source code into a structural model using its own parser (Clarpse), not regex or text matching.

How does the AI work? Is it just an LLM wrapper?

No. Striff uses a neurosymbolic pipeline: deterministic structural analysis (8 detectors for cycles, coupling, boundaries, complexity) combined with a graph neural network trained on 129 well-architected codebases (AUC 0.91). The AI explains findings — it doesn't invent them. Every note is grounded in architectural facts.

Does it slow down my PR pipeline?

No. Striff runs as a GitHub Check, typically completing in under 30 seconds. It doesn't block merges unless configured to. The structural diff and review notes appear in the Checks tab.

What's the difference between the free extension and the GitHub App?

The browser extension is free and works on public repositories — you view structural diffs directly on GitHub PR pages. The GitHub App adds automated checks on every private PR, shared team workflow, and billing management. Plans start at $29/month.

Is my code sent to a third party?

Striff parses your code to extract structural information (components and relationships). Source code is processed in memory and not persisted. For enterprise needs, self-hosted deployment is available.

Architecture review for every pull request.

Striff works alongside your existing review tools. Install free for public repos — it runs on your next PR.