VEX vs SBOM: component inventory and exploitability context
An SBOM lists the components in a software product. VEX communicates whether a known vulnerability affects that specific product and why. They answer different questions and work together: inventory supports matching; VEX adds a reasoned product-level status for a finding.
By the Vellaci teamPublished Updated 8 min readOperational guidance, not legal advice
Key facts
- SBOM asks
- Which components and dependency relationships are in this product version?
- VEX asks
- Does this vulnerability affect this product, and what is the status and rationale?
- Relationship
- Use component and version matching to find candidate vulnerabilities; review product context before issuing a VEX statement.
- Formats
- CycloneDX can represent SBOM and VEX data, in one document or two; OpenVEX and CSAF also support VEX use cases.
- Statuses
- Not affected, affected, fixed, under investigation (CISA minimum requirements); CycloneDX expresses them as analysis states such as not_affected, exploitable, resolved and in_triage
- CRA caveat
- VEX can support vulnerability-handling evidence. It is not itself a CRA certificate or a substitute for the manufacturer's assessment.
What is the difference between VEX and an SBOM?
An SBOM is an inventory of software components and their relationships for a product or release. A VEX statement is an advisory about the impact of a known vulnerability on a particular product. An SBOM helps identify where a vulnerable package may be present; VEX records the manufacturer's product-specific status and reasoning after review.
| Aspect | SBOM | VEX |
|---|---|---|
| Primary question | What components are included? | Is this vulnerability exploitable or otherwise affecting this product? |
| Main subject | Components, versions, suppliers and dependency relationships | A vulnerability in a named product or product version |
| Typical inputs | Build, package manager, image or firmware inventory | A vulnerability finding plus product configuration, code and reachability context |
| Typical output | Machine-readable inventory | Status, justification, action and supporting detail in a supported VEX format |
| What it does not prove | That every listed component is vulnerable or exploitable | That all product risks are assessed or that the product conforms to law |
How they work together
A component matcher compares package identity and version information from an SBOM with vulnerability advisories. That produces candidate findings, not a final product-impact determination. A reviewer checks the affected version range, product configuration, vulnerable code presence or reachability, mitigations and available fixes. The reviewer then records the decision, evidence and any action. Where useful, the decision is expressed as a VEX statement for customers or downstream systems.
- Generate and preserve the SBOM for the exact product release.
- Match component identifiers and versions against vulnerability data.
- Verify the affected range and investigate the finding in product context.
- Record a human-reviewed status, rationale, evidence, owner and remediation action.
- Publish or share VEX in a supported format when it helps downstream users.
Example: one component, one CVE, one VEX statement
The release SBOM for a gateway firmware lists golang.org/x/net at v0.15.0 with its package URL. CVE-2023-44487, the HTTP/2 'Rapid Reset' vulnerability, affects that module below v0.17.0, so a component matcher raises a finding. The SBOM excerpt below is the evidence that the component shipped; it says nothing about exploitability.
{
"bomFormat": "CycloneDX",
"specVersion": "1.6",
"serialNumber": "urn:uuid:3e671687-395b-41f5-a30f-a58921a69b79",
"version": 1,
"metadata": {
"timestamp": "2026-10-20T08:00:00Z",
"component": { "bom-ref": "acme-gateway@4.2.0", "type": "firmware", "name": "acme-gateway", "version": "4.2.0" }
},
"components": [
{
"bom-ref": "pkg:golang/golang.org/x/net@v0.15.0",
"type": "library",
"name": "golang.org/x/net",
"version": "v0.15.0",
"purl": "pkg:golang/golang.org/x/net@v0.15.0"
}
]
}The VEX statement for the same finding
The product team establishes that the gateway's management interface serves HTTP/1.1 only: the HTTP/2 server code in the module is linked but never started, and no setting exposes it. The VEX document records that conclusion against the exact component in the exact release, using a BOM-Link (urn:cdx:serial/version#bom-ref) back to the SBOM. If a later release enables HTTP/2, the statement no longer holds and a new one is issued.
{
"bomFormat": "CycloneDX",
"specVersion": "1.6",
"version": 1,
"metadata": { "timestamp": "2026-10-20T09:00:00Z" },
"vulnerabilities": [
{
"id": "CVE-2023-44487",
"source": { "name": "NVD", "url": "https://nvd.nist.gov/vuln/detail/CVE-2023-44487" },
"analysis": {
"state": "not_affected",
"justification": "code_not_reachable",
"detail": "The HTTP/2 server in golang.org/x/net is linked but never started; the management interface serves HTTP/1.1 only and no setting enables HTTP/2.",
"firstIssued": "2026-10-20T09:00:00Z",
"lastUpdated": "2026-10-20T09:00:00Z"
},
"affects": [
{ "ref": "urn:cdx:3e671687-395b-41f5-a30f-a58921a69b79/1#pkg:golang/golang.org/x/net@v0.15.0" }
]
}
]
}Worked example
A release SBOM identifies a specific version of a logging library. An advisory indicates that a version range may contain a vulnerability. The match creates a finding for review. The product team checks whether the affected code is present, reachable and configured in the shipped product, and whether a fixed version is available. The SBOM remains the component inventory; the resulting VEX statement records the product-specific status and the evidence behind it. A new product release or new evidence may require the VEX assessment to be revisited.
Status terms depend on the format
CISA's minimum requirements for VEX define four statuses — not affected, affected, fixed and under investigation — and OpenVEX and CSAF follow that model. CycloneDX records the same decision as an analysis state on the vulnerability, with its own vocabulary: there is no 'affected' state, the closest being exploitable. Map deliberately, validate against the chosen format's schema, and keep the justification and update history with every statement.
| Meaning | CISA / OpenVEX status | CSAF 2.0 product status | CycloneDX analysis.state |
|---|---|---|---|
| The product is not affected | not_affected (justification required) | known_not_affected (with a flag or impact statement) | not_affected (with a justification) |
| The product is affected | affected (action statement expected) | known_affected | exploitable |
| A fix is available in these versions | fixed | fixed / first_fixed | resolved (resolved_with_pedigree when the fix is shown in pedigree) |
| Still being assessed | under_investigation | under_investigation | in_triage |
| The match was wrong | — (not part of the CISA model) | — | false_positive |
Justifications for 'not affected'
A not-affected status without a reason is an assertion. CISA's VEX status justifications, reused by OpenVEX and as CSAF flags, name five reasons; CycloneDX has a finer list of nine. The right-hand column is our closest mapping, not an official crosswalk. Use the most specific justification your evidence supports and explain it in the detail field.
| CISA / OpenVEX / CSAF justification | Meaning | Closest CycloneDX justification |
|---|---|---|
| component_not_present | The vulnerable component is not in the product | code_not_present, or false_positive when the component was matched in error |
| vulnerable_code_not_present | The component is present but the vulnerable code was excluded at build time | code_not_present |
| vulnerable_code_not_in_execute_path | The vulnerable code is present but never executed | code_not_reachable |
| vulnerable_code_cannot_be_controlled_by_adversary | The code runs, but an attacker cannot reach or control its inputs | requires_configuration, requires_dependency, requires_environment |
| inline_mitigations_already_exist | Built-in protections that users cannot disable prevent exploitation | protected_by_compiler, protected_at_runtime, protected_at_perimeter, protected_by_mitigating_control |
How VEX relates to the Cyber Resilience Act
Annex I Part II of the CRA requires manufacturers to identify and document components and vulnerabilities and to operate vulnerability-handling processes. VEX can help communicate product-specific findings and preserve part of the decision evidence. The regulation does not turn a VEX document into proof of conformity: the manufacturer remains responsible for the applicable assessment, remediation and technical documentation.
Official requirement, interpretation and practice
Official requirement: no article of Regulation (EU) 2024/2847 names VEX. The text asks manufacturers to identify and document components and vulnerabilities, including an SBOM (Annex I Part II(1)), to systematically document vulnerabilities they become aware of (Article 13(7)), and to publicly disclose information about fixed vulnerabilities once a security update is available (Annex I Part II(4)).
Interpretation: customers, integrators and market surveillance authorities will ask whether a specific CVE affects a specific version. A per-version VEX statement with a justification is a direct, machine-readable answer to that question, and it keeps the reasoning attached to the decision instead of in an e-mail thread. It is one practical way to evidence the documentation the regulation does require.
| Statement | Status in the CRA | What it supports |
|---|---|---|
| SBOM per version | Required (Annex I Part II(1)) | Identifying components and matching vulnerabilities |
| Documented vulnerability decisions | Required (Article 13(7)) | Showing what was known, decided and why |
| Disclosure of fixed vulnerabilities | Required once an update is available, with justified delays (Annex I Part II(4)) | Users remediating |
| VEX document | Not named; a common format for the above | Machine-readable, product-specific status and justification |
What a VEX decision should keep as evidence
A status alone is weak evidence. For each statement, keep the product and exact version, the vulnerability identifier and its source, the status, the justification code where the status is not affected, a short free-text detail explaining the reasoning, the reviewer and the time of the decision, and the action or fixed version where the status is affected or fixed. Keep the history: when new evidence changes a decision, record a new statement rather than editing the old one, so the record still shows what was believed at the time a customer or an authority was told.
Revisit statements when a new release ships, when the advisory's affected range changes, when exploitation evidence appears (for example a CISA KEV listing), or when a configuration change alters reachability. A not-affected statement for version 2.3 says nothing about version 2.4 until someone reviews it.
Frequently asked questions
- Does an SBOM say whether my product is vulnerable?
- No. It records components and relationships. Vulnerability matching produces candidate findings; product configuration, affected versions and other context still need review.
- Does VEX replace an SBOM?
- No. VEX communicates vulnerability impact for a product. It does not provide the full component inventory that supports repeatable matching and traceability.
- Can VEX say a component is not affected?
- A VEX statement can record a product-specific not-affected status with a justification supported by evidence. Use the status terms and structure defined by the chosen VEX format.
- Does the CRA require a VEX file?
- The CRA requires vulnerability handling and component/vulnerability documentation, including an SBOM covering at least top-level dependencies. VEX is a useful communication and evidence format; the regulation does not prescribe a single VEX format as a conformity certificate.
- Can SBOM and VEX be in the same file?
- In CycloneDX, yes: a BOM can carry components and a vulnerabilities array with analysis states. Many manufacturers still publish VEX as a separate document that references the SBOM by BOM-Link, because VEX changes whenever a new advisory is assessed while the SBOM of a shipped release should not change.
- Which VEX format should I use?
- Use the format your customers' tooling consumes. CycloneDX VEX fits teams already producing CycloneDX SBOMs; CSAF 2.0 is common for vendor security advisories; OpenVEX is a minimal format built on the CISA model. The statuses map onto each other, so the decision record matters more than the serialisation.
- What does 'not_affected' with 'code_not_reachable' mean?
- That the vulnerable code is in the product but cannot be executed as the product uses it — the CycloneDX counterpart of CISA's vulnerable_code_not_in_execute_path. Keep the evidence, such as configuration, build flags or call-graph analysis, and revisit the statement when the product changes.
From requirement to evidence
What the obligations in this guide mean for an engineering, security or compliance team, the Vellaci workflow that supports each one, and the records it leaves. The full map for every CRA obligation is on the CRA compliance software page.
How to read this. Official requirement summarises the text of Regulation (EU) 2024/2847 with its reference. What your team does is Vellaci’s operational interpretation, not legal advice. Vellaci workflow and Evidence describe what the product does today. Methodology · reviewed 2026-10-05.
- Official requirement
- In relation to the risks posed, address and remediate vulnerabilities without delay, including by providing security updates; where technically feasible, provide security updates separately from functionality updates.
- What your team does (interpretation)
- Every matched vulnerability gets a human decision — affected, not affected, fixed, accepted — with a reason, an owner and, if affected, a remediation target and the version that fixes it.
- Vellaci workflow
- Triage with mandatory reasons, explainable priority (CVSS, EPSS, KEV, exposure), remediation owner and due date, Jira and Linear ticket links. How it works
- Evidence produced
- Decision history per finding with reviewer and timestamp
- Remediation owner, due date and fixed version
- Linked tickets
- Official requirement
- Systematically document, proportionately to the nature and the cybersecurity risks, relevant cybersecurity aspects of the product, including vulnerabilities the manufacturer becomes aware of and relevant information provided by third parties, and update the risk assessment where applicable.
- What your team does (interpretation)
- Keep the reasoning, not just the status. A 'not affected' without a justification is not documentation; neither is a status that was silently overwritten.
- Vellaci workflow
- Append-only decision history and hash-chained audit log; VEX states with CycloneDX justification codes per product version. How it works
- Evidence produced
- Per-product VEX statements with justification
- Audit log entries with before and after values
- CycloneDX VEX export per version
- Official requirement
- Once a security update has been made available, share and publicly disclose information about fixed vulnerabilities, including a description, the affected products, impact, severity and information helping users remediate; publication may be delayed in duly justified cases.
- What your team does (interpretation)
- Tie each advisory to the finding and the fix release, and record when and where it was published — or why publication was delayed.
- Vellaci workflow
- Fixed versions recorded on findings; CycloneDX VEX export for customers and downstream integrators. How it works
- Evidence produced
- Fixed version linked to the finding
- Exported VEX document
- Publication or delay rationale