Every modern application is assembled from open-source parts, and each one arrives with two things you did not write: a license that may put obligations on your own code, and a version history that may contain known security holes. OSSScan exists to give you clear insight into both: what your open-source licenses require of you, and which of the CVEs buried in your dependency tree can actually hurt you.
“But isn't Dependabot enough?”
It's a fair question, especially for the security half. If you host on GitHub you already have Dependabot, it's free, and it's genuinely good at one job: watching your dependencies around the clock and raising an alert the moment a new CVE is published against something you use. As an early-warning system, it's excellent. So why pay for anything more?
Because being alerted and knowing what to do are two very different things, and this is where Dependabot's one job becomes its one shortcoming.
An alarm is not an answer
A Dependabot alert isn't a fact about your application; it's a hypothesis. It matches a package and version against an advisory database and raises a flag. What it can't tell you is whether the broken code is reachable from your app at all. Is the vulnerable function even called? Can attacker-controlled input reach it? Does the package ship to production or only run in your tests? So it does the only safe thing it can and flags everything. The result is a laundry list: dozens or hundreds of red rows, most of which don't matter for your situation, and no way to tell which few do. GitHub does offer reachability analysis to help here, tracing whether vulnerable code is actually called, but on private repositories it is a paid add-on (part of GitHub Advanced Security, now GitHub Code Security), and like any static-analysis approach its usefulness depends on language coverage and what the analyzer can see.
That's the noise. Working through it by hand, opening the code and answering the reachability question one advisory at a time, is exactly the expensive, tedious work OSSScan is built to take off your plate. It hands each finding to the AI coding agent you already use, which reads your code and comes back with a verdict: exploitable, not exploitable, or needs a closer look, with its reasoning and a human sign-off on top. Instead of a laundry list, you get a short list of what actually matters.
Better together: OSS as a test pass
Here's where the two tools stop competing and start compounding. Treat your open-source review the way you treat your test suite: a test pass you run deliberately, at a moment you choose, that certifies the state of the code. That's OSSScan's job. Then, in between passes, let Dependabot do what it does best: keep watch and alert you to genuinely new CVEs published since your last pass.
OSSScan
The periodic test pass
A deep, AI-assisted investigation you run and control, covering reachability, license compliance, and a human sign-off, until the scan is clean.
Dependabot
Real-time monitoring in between
The always-on alarm that flags new CVEs as they land, free on GitHub, so nothing slips through between your test passes.
What makes it work: Update Dependabot
The pairing only holds together because OSSScan can hand its conclusions back to Dependabot. That's the Update Dependabot feature. When a test pass is clean, one click writes your investigated verdicts to GitHub, dismissing the matching Dependabot alerts, each with a reason and a comment recording that OSSScan cleared it, when, and why.
That's the piece that makes real-time monitoring livable. Without it, Dependabot would keep re-raising every alert you already investigated, and the laundry list would grow forever. With it, the findings you've cleared go quiet, so between test passes the only alerts Dependabot shows you are the genuinely new ones: a manageable stream instead of a wall. You get to keep continuous CVE monitoring switched on and keep it quiet.
Careful by design. Update Dependabot previews every change before writing anything, dismisses only findings you've cleared, and never silently overrides an alert a human dismissed by hand. It's strictly one-way: OSSScan writes its verdicts to Dependabot and leaves the live view to Dependabot. And you hold the keys: you mint a short-lived GitHub token scoped to just Dependabot alerts, paste it once, and revoke it after. It can't read your code or touch anything else.
Run passes on your cadence
Put it together and you get real-time CVE monitoring running continuously, punctuated by deep OSSScan test passes on whatever schedule fits your project: every release, monthly, quarterly, once a year, your call. A fast-moving service might run a pass each sprint; a stable library, once a quarter. Between passes, Dependabot keeps watch and OSSScan keeps it quiet.
And that's why the licensing fits
This cadence is exactly why OSSScan is sold as a short-term license rather than a standing subscription. You're not running it year-round. You're running a focused pass now and then, so you pay for a focused pass now and then. Buy a short pass when it's time for a deep review, run the full investigation, push your verdicts to Dependabot, and let the free layer carry you to the next one. You pay for depth when you need depth, and nothing the rest of the year. That's the whole idea: better together.
← Back to the OSSScan Blog