Software supply chain security: managing the risk in your dependencies

Line drawing of parcels arriving on conveyors from several directions into one box, one parcel checked and outlined in blue

Most of the code you ship was written by someone else. A step-by-step guide to knowing what is in your software, choosing what to trust, protecting the build and reacting fast when a component is compromised.

In short

Software supply chain security is the work of controlling risk from everything that goes into your software: open-source packages, vendor SDKs, build tools and pipelines. Start with an SBOM for every build and continuous vulnerability monitoring, then add rules for choosing and updating dependencies, a hardened and traceable build, and a rehearsed response for the day a key component is hit.

Log4Shell in December 2021 showed how one logging library could expose thousands of products at once. The xz Utils backdoor in 2024 showed something worse: a trusted maintainer role used to plant malicious code. This guide is written by Devies, a Gothenburg-based product and engineering consultancy, for CTOs and engineering leads who want a practical approach to supply chain security without stopping developers from using open source.

What is software supply chain security?

Software supply chain security is the protection of the integrity and trustworthiness of every component, tool and process that turns source code into the software your customers run.

Key terms:

  • A dependency is any external code your software uses, directly or through other packages (transitive dependencies).
  • An SBOM (software bill of materials) is a machine-readable inventory of the components and versions in a piece of software.
  • Provenance is verifiable information about where, how and from what sources an artefact was built.
  • A software supply chain attack is a compromise of a component, tool or distribution channel in order to reach the organisations that use it.

The risks fall into four groups, and each needs a different control:

Risk

Example

Main control

Known vulnerabilities in components

Example

An outdated library with a published CVE

Main control

SBOM plus continuous scanning

Malicious or hijacked packages

Example

Typosquatted names, a compromised maintainer account

Main control

Package vetting, pinning, private registry

Compromised build or pipeline

Example

Stolen CI credentials, tampered build scripts

Main control

Hardened CI, short-lived credentials, provenance

Abandoned components

Example

No maintainer to fix a new flaw

Main control

Health checks before adoption, replacement plans

The CISA alert on the xz Utils compromise (CVE-2024-3094) is worth reading as a case study of the second and fourth rows combined: a long-running effort to gain maintainer trust in a widely used but thinly staffed project.

How do you secure your supply chain, step by step?

Secure the supply chain in a fixed order: first know what you have, then decide what you accept, then protect how you build, and finally prepare to respond.

Four connected platforms with a list, a gate, a sealed factory and a bell, the sealed factory in blue
  1. Generate an SBOM in every build. Use CycloneDX or SPDX, store it with the release artefact and keep it for every version you support. See what an SBOM is and why the CRA requires one.
  2. Monitor continuously, not only at build time. Feed SBOMs into a tool that alerts when a new vulnerability affects a shipped version. Prioritise with the CISA Known Exploited Vulnerabilities catalogue and real reachability, not CVSS scores alone.
  3. Set adoption rules for new dependencies. Before adding a package, check maintenance activity, number of maintainers, licence and security practices. OpenSSF Scorecard automates many of these checks for open-source projects.
  4. Pin and lock versions. Commit lock files, pin exact versions and verify checksums, so builds are reproducible and a new malicious release does not slip in silently.
  5. Use a private registry or proxy. Mirror approved packages, block known bad ones and reduce the risk of dependency confusion between internal and public package names.
  6. Update deliberately and often. Automated update pull requests in small batches are safer than rare big-bang upgrades. Review changes for critical packages.
  7. Harden the build pipeline. Use short-lived credentials, least-privilege tokens, protected branches, required reviews and isolated build runners. Keep secrets out of code and logs; see secrets management.
  8. Produce provenance and sign artefacts. The SLSA framework defines levels of build integrity, from documented builds to hardened, tamper-resistant build platforms with signed provenance.
  9. Prepare a component incident playbook. When the next Log4Shell lands: who checks the SBOMs, who decides on emergency releases, how customers are informed.

How do you handle vendor and commercial components?

Treat vendor SDKs, commercial libraries and SaaS integrations as part of the same supply chain, because they reach your product with the same privileges as open source and often with less visibility.

Practical requirements to put in contracts and onboarding:

  • An SBOM or component list for delivered software, updated with each release.
  • A vulnerability disclosure contact and a committed time to notify you of security issues.
  • A defined support period and update policy.
  • Evidence of secure development, for example alignment with the NIST Secure Software Development Framework or an equivalent.

For organisations covered by NIS2 and the Swedish Cybersecurity Act (cybersäkerhetslagen), supply chain security is one of the explicitly required risk-management measures, so supplier requirements also serve compliance. Manufacturers under the Cyber Resilience Act must exercise due diligence when integrating third-party components and, as of October 2026, report actively exploited vulnerabilities in their products.

Who owns supply chain security?

Supply chain security works when ownership is split clearly: engineering owns the pipeline and dependency choices, security owns the policy and monitoring, and product management owns the support period and the customer communication.

Common gaps appear between those roles. A vulnerability alert lands with a security team that does not know which product versions are in the field, while the product team does not see the alert at all. Agree on one triage queue, severity-based deadlines and an exception process with expiry dates. Report a small set of metrics monthly: share of builds with an SBOM, open critical findings in shipped versions, median time to patch, and number of dependencies without an active maintainer.

Devies saw this pattern in its work with Zenseact on safety-critical automotive software, where a vulnerability and dependency tracking solution based on SBOMs was hosted and integrated into CI/CD as part of daily development rather than as a separate compliance process.

Where do scanners fit?

Software composition analysis is the scanner that does most of the work in supply chain security, but it covers only known vulnerabilities in declared components, not tampered builds or malicious new releases.

Combine it with static analysis of your own code and dynamic testing of the running system; the trade-offs are covered in SAST, DAST or SCA. Then add the controls a scanner cannot provide: pinning, provenance, adoption rules and the response playbook.

Devies

Want a second opinion?

Devies, a Gothenburg-based product and engineering consultancy, works on questions like this with Nordic product and IT teams. Tell us where you are and we will suggest a sensible first step.

Talk to Devies

Frequently asked questions

What is a software supply chain attack?

A software supply chain attack compromises something your software depends on, such as an open-source package, a build tool, a vendor update or a CI pipeline, in order to reach the organisations that use it. Examples include malicious code inserted into a library by a compromised or rogue maintainer, typosquatted package names and tampered build servers. One successful attack can reach thousands of downstream users at once.

Is an SBOM enough for supply chain security?

No. An SBOM tells you what is inside your software, which is the foundation for finding affected versions quickly. It does not stop a malicious package from being added, prove that a build was not tampered with or tell you whether a component is still maintained. Combine it with continuous monitoring, adoption rules, version pinning, a hardened pipeline, signed provenance and a response playbook.

What is SLSA?

SLSA (Supply-chain Levels for Software Artifacts) is an open framework, maintained under the OpenSSF, that defines increasing levels of build integrity. Lower levels require documented, automated builds with provenance; higher levels require hardened build platforms that make provenance hard to forge. Teams use it as a roadmap for protecting the path from source code to released artefact, and to communicate build assurances to customers.

How often should we update dependencies?

Continuously and in small steps. Automated update pull requests reviewed weekly keep each change small and easy to test, and they make emergency patches much faster because you are never many versions behind. Security fixes for known exploited vulnerabilities should be applied within days. For critical packages, review release notes and changes before merging, since a new release can also introduce risk.

Sources

Here’s more

Join Our Newsletter

We handle your data according to our privacy policy.

Shopping Cart

We use cookies

We use cookies to improve your experience on this website. You may choose which types of cookies to allow and change your preferences at any time. Disabling cookies may impact your experience on this website. You can learn more by viewing our Cookie Policy.

Check out

Contact Form text

We will contact you text

We handle your data according to our privacy policy.