See where your stack is falling behind.
Vibgrate scans your repository on your machine. It shows how far your stack has fallen behind and finds dependencies with known vulnerabilities. Then it gives your coding agent focused context from the code you actually run.
npx @vibgrate/cli scanno install·Nothing is installed globally — ideal for CI or a one-off scan.
Free · no account · results in about 60 seconds · raw source is not uploaded
A replay of the actual CLI against our test repos — nothing executes in your browser.
Don’t take our word for it.
Vibgrate is built in the open. Inspect the code, review the methodology, follow the releases, and verify the results for yourself.
Inspect the CLI, the scanner implementation, and the public repositories it runs against.
Review the scoring model, its assumptions, its evidence sources, and its stated limitations.
See the datasets, methodology and results behind every performance claim we make.
Follow every improvement, fix and capability added to the platform.
Your stack moves away from current, supported versions.
Runtimes pass their end-of-life date. Frameworks fall major versions behind. Dependencies stop getting security fixes. Each one is small on its own. Most teams find out during the upgrade that hurts, or when a CVE lands on a package nobody has touched in two years.
Coding agents can make the problem worse. Without the right context, they can suggest APIs or patterns that do not match the versions in your project. Vibgrate measures that gap, then gives the agent the versions and code you actually have.
Three things, from one free CLI.
Version-correct documentation, a map of your code, and a score for how far you have fallen behind. Each one works on its own.
Docs for the versions in your lockfile.
Give your coding assistant documentation for the versions you actually use, together with focused context from your repository. It works locally, with no account required.
Version-correct contextCode mapMaps your repository once.
Your coding agent can use that map instead of repeatedly searching the same files, so it spends fewer tokens on the same work.
How the map worksDriftScoreOne number for how far behind you are.
See how far your runtime, frameworks, and dependencies have fallen behind. Vibgrate gives the project a 0–100 DriftScore and shows what to upgrade next.
DriftScore, explainedIn our published benchmarks, focused context used about 29% fewer tokens for the same tasks on a ~300-file repository. See token benchmarks
~29% fewer tokens, same results
The same coding tasks, the same model, with and without the Vibgrate code map — on a ~300-file codebase. We count only the tasks both runs solved, so it is fewer tokens for the same result, never “cheaper because it gave up”.
Measured with a frontier coding model. On tiny repositories grep already finds the code and the map costs more — we publish that part of the curve too. See the full benchmark
Measure it. Rank it. Ship it.
See a 0–100 DriftScore for each project — and across your whole portfolio.
How the DriftScore worksRankAn upgrade plan ordered by effort, risk, and the cost of waiting.
Tour Vibgrate CloudShipHand the work to the coding agent you already use. Vibgrate checks the result afterward.
Coding-tool integrationsDescribe the change. Read the diff. Decide.
VG Code maps your repository once, so the model works with focused context instead of repeatedly searching your code. Every change arrives as a real diff. Nothing changes until you say so.
Because the model receives less context, hosted runs use fewer tokens — and cost less at the same model rate. With Vibgrate Relay, the model price is visible before you run it.
You read a real diff, then approve it. Nothing changes until you say so.
The same approvals and behavior everywhere.
Works from your real code
VG Code uses the structure already mapped from your repository. If a proposed change depends on something that is not there, it stops before changing a file.
Real diffs you approve
Every edit and command waits for your approval. You can review the plan before anything changes, and undo an approved change afterward.
See what else will be affected
Before a change, VG Code shows the related code and tests it could affect. Afterward, it runs your own checks against the result.
Local or hosted — same experience
Run a model on your machine or use hosted models through Relay. VG Code recommends a local model that will run well on your machine, with no separate setup. You can still choose any model you prefer.
VG Code is included with Vibgrate CLI and Vibgrate for VS Code. No account required. Run models locally, use your own provider, or use hosted models through Relay.
The overview your team shares
See DriftScore, critical risks, and project health across your repositories. The example below uses sample data.
Example data · no account, no card · push one scan to replace it with your own
Explore Vibgrate
Use the parts you need. They work together, but each one stands on its own.
Vibgrate CLI
Scan any repository for drift and known vulnerabilities. Free, and it runs on your machine.
Learn moreVG Code
Describe a change and read the diff before anything happens.
Learn moreVibgrate for VS Code
The score and the coding agent inside your editor.
Learn moreVibgrate Cloud
The shared view of drift, risk, and upgrade work.
Learn moreVibgrate Relay
Hosted models for VG Code, with the price visible before you run.
Learn moreVibgrate Evidence
Signed evidence of which shipped products contain a vulnerability.
Learn moreUseful from one repository to the whole portfolio.
Developers
Give your coding assistant the right versions, and see what an upgrade could affect before you open the PR.
Tech Leads
See repository health, structural drift, and the upgrades worth doing next. Analysis runs on your machine.
Platform Engineers
Set drift and vulnerability limits in CI. Export SARIF when you need it. Run everything offline.
Security Engineers
See which vulnerabilities matter, and where outdated dependencies are creating risk. Export SBOMs and evidence for CRA work.
Engineering Managers
See which teams are drifting, and what one sprint of upgrades buys you.
CTOs
Is your estate modernizing or drifting? Portfolio risk on one board-ready page.
Independent by design.
Vibgrate measures the drift, ranks the work, and checks the result. You choose where the work runs — on your machine, in your cloud or CI, using the models and coding agents you prefer.
Free to scan. Pay when your team needs governance.
Scanning is free. Paid plans add shared governance and are priced by project, with unlimited seats. New customers get 25% off their first year.
Get your DriftScore in minutes.
Install the CLI and run one scan. See how far your project has drifted, then give your coding agent the context it needs to work with the code you actually have.
npx @vibgrate/cli scanno install·Nothing is installed globally — ideal for CI or a one-off scan.
No account · results in about 60 seconds · runs on your machine, raw source is not uploaded