Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Software composition analysis identifies the open-source components in an application, matches them against known vulnerabilities and licences, and tracks them as they change.

What Is Software Composition Analysis?
Most of the code in a modern application was written by someone else. A typical JavaScript service resolves well over a thousand packages, a Java service a few hundred, and almost none of them were chosen deliberately: they arrived as dependencies of dependencies. Software composition analysis is the practice of knowing exactly what those components are, which published vulnerabilities apply to them, what licences they carry, and when any of that changes.
A scanner builds an inventory first. The reliable method reads the resolved dependency graph, from a lockfile, a build output, or a compiled artefact, rather than the declared manifest, because the manifest states intent and the lockfile states fact. Container images are analysed layer by layer, which picks up operating system packages alongside application libraries.
Each component is then matched against vulnerability data: the National Vulnerability Database, the OSV database that aggregates ecosystem-specific advisories, GitHub’s advisory database, and vendor feeds. Matching is the part that goes wrong. Version strings differ between ecosystems, vendored and repackaged code hides its origin, and advisories sometimes name a product rather than a package. Expect both missed components and false matches, and expect to verify rather than to accept.
The dependency you installed is rarely the one that is vulnerable. Fixing a transitive issue means persuading the direct dependency to accept a newer version of its own, and when it will not, the options are a version override, a fork, or replacement. Scanners that report only direct dependencies miss most of the real exposure.
Reachability analysis is what makes the output manageable. A vulnerability in a function your code never calls still deserves patching, and it is not an emergency. Tools that trace call paths from your code into the dependency can separate the two, which commonly reduces the urgent list by an order of magnitude. Treat the result as prioritisation rather than as permission to ignore, since a future change may make the unreachable path reachable.
Log4Shell, CVE-2021-44228, was published on 9 December 2021. A logging library present in an enormous share of Java applications evaluated JNDI lookups found inside logged strings, so writing a crafted string into any logged field produced remote code execution. The exploit was trivial and the hardest part of the response, for nearly every organisation, was answering the question of where the library was installed. Teams with a dependency inventory answered in hours; teams without one spent weeks.
The xz-utils backdoor, CVE-2024-3094, was disclosed on 29 March 2024. A maintainer account built up over years inserted obfuscated code into release tarballs of a widely used compression library, targeting SSH authentication on systems that linked it indirectly. No scanner would have flagged it before disclosure, because the component was not known to be vulnerable; what mattered was knowing quickly whether the affected versions were present. That is the argument for an inventory independent of the argument for vulnerability matching.
The NCSC’s supply chain security guidance expects organisations to understand and manage the components they depend on, and to hold suppliers to the same standard. The EU Cyber Resilience Act, which entered into force on 10 December 2024, will require manufacturers placing products with digital elements on the EU market to maintain a software bill of materials and to handle vulnerabilities throughout the product lifetime, with the main obligations applying from December 2027. UK organisations selling into the EU are in scope. Since this control also addresses dependency confusion and other build-time manipulation, it is a standard area of review during our penetration test services.
They answer different questions. A network vulnerability scanner probes running hosts and services for known weaknesses from the outside. Software composition analysis reads your dependency graph and matches it against advisories, which finds libraries embedded in an application that no network probe would reveal. Most programmes need both, since neither sees what the other does.
No, and attempting it exhausts the team. Prioritise by whether the vulnerable code is reachable from your own, whether the component is exposed to untrusted input, whether exploitation is known to be occurring, and what the realistic impact would be. Record the reasoning for anything deferred, with an owner and a review date, so the decision is visible later.
It determines whether your code can actually invoke the vulnerable function in a dependency, rather than merely including the package. Tools trace call paths from your entry points into library internals. The distinction matters because a large majority of reported issues sit in code paths an application never executes, and separating those from the genuinely exposed ones makes the backlog tractable.
It solves the inventory half. An SBOM tells you what is in a release, which is exactly what organisations could not answer during Log4Shell and the xz-utils backdoor. It does not tell you what is vulnerable today, so it has to be checked continuously against advisory data as new issues are published against components you already shipped.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.