Function reference
A complete map of search, reports, national guidance, KEV, Patch Intelligence, change tracking, exports and API access.
Open guide ↗How-to guides
Short workflows for common tasks, from one CVE lookup to a copy-ready multi-CVE assessment.
Open guide ↗Data and limitations
How sources, scoring, remediation evidence, freshness, privacy and uncertainty are handled.
Choose the task you are trying to complete
Search a CVE, EUVD, BSI WID-SEC or supported advisory identifier, then open the full report.
Show me how ↗Compare a listPaste identifiers or upload a TXT or CSV file, then copy the email-ready assessment table.
Show me how ↗Prioritise urgent workUse Known Exploited, EPSS, Patch Tuesday and Recent Changes without confusing prediction with confirmed exploitation.
Show me how ↗Match an SBOM or VEX fileCheck CycloneDX or SPDX components against the catalogue, optionally apply VEX statements, and review ambiguous matches.
Show me how ↗Understand a resultReview which authority supplied the claim, whether product versions are mapped and whether a fix is verified.
Explain the evidence ↗How the site is organised
- CVE Intelligence
- The searchable working view. Use it to find identifiers, filter the catalogue, inspect evidence and open a complete report.
- CVE archive
- A crawlable year-by-year directory of canonical reports. It is useful when browsing rather than searching, and when linking to a stable report URL.
- Known Exploited
- A focused list of records currently backed by CISA KEV evidence. It prioritises confirmed exploitation, not technical severity alone.
- Patch Tuesday
- Vendor bulletin and release-cycle intelligence. Use it to move from a monthly advisory to its linked CVEs, products, source material and remediation checklist.
- Recent changes
- A feed of material field changes after publication. It suppresses routine refreshes so that severity, exploitation, product and remediation changes stand out.
- Multi-CVE lookup
- A list workflow for pasted identifiers or TXT and CSV uploads, with linked results and reusable email, spreadsheet and file exports.
- Inventory applicability
- A separate private matching workspace for supported CycloneDX and SPDX files, optional VEX statements, and careful review of ambiguous or unmatched components.
A practical evidence checklist
- Confirm the identifier.
Check that the CVE, EUVD or advisory identifier resolves to the record you intended and that the title describes the expected product.
- Separate severity from exploitation.
Read CVSS, EPSS, public exploit references and KEV as different signals. None of them replaces the others.
- Match the exact product branch.
Use product-aware affected and fixed rows. Do not transfer a version range from one product, edition, operating system or cloud service to another.
- Open the cited source.
Use the vendor or authority link to verify the latest revision, prerequisites, workarounds, restart requirements and exceptions.
- Apply your environment context.
Exposure, installed configuration, compensating controls, service criticality and change risk belong to your own operational decision.
Questions and detailed answers
Why can CVSS, EPSS and KEV point in different directions?
They measure different things and should not be read as competing versions of one risk score.
- CVSS describes technical severity. It summarises characteristics such as attack vector, required privileges, user interaction and potential impact. It does not know whether your organisation uses the affected product or whether attackers are currently exploiting it.
- EPSS is a probability forecast. It estimates the likelihood that a published CVE will be exploited in the near term based on observed signals. A low probability is not proof that exploitation is impossible, and a high probability is not confirmation that exploitation has occurred.
- KEV records confirmed exploitation. A CISA KEV listing means CISA has evidence of exploitation in the wild. Absence from KEV means that this particular confirmation is not present, not that the CVE is safe.
For example, a medium-severity vulnerability can appear in KEV because attackers value its reliability or reach, while a critical vulnerability can have a low EPSS value because it is difficult to operationalise. Start with confirmed exploitation, then combine technical impact, exposure, affected product and your own business context.
Does “Patch available” mean every affected product has a fix?
No. It means an authoritative patch reference, fixed product state or vendor-authored solution was found somewhere in the retained evidence. Always read the product-specific affected and fixed rows before scheduling a change.
- Different release branches: a vendor may fix Product A 5.0 in 5.0.6 and Product A 4.4 in 4.4.9. “Patch available” does not mean that every installation should move to one shared version.
- Different products: one advisory may cover an appliance, cloud service and platform service. The appliance may require an upgrade, the cloud service may already be corrected by the vendor, and one older branch may have no fix.
- Partial structured evidence: BlackTree may find an authoritative vendor update link while the CVE record itself contains no fixed-version field. The report then links to the advisory but keeps the fixed version unverified until a product-specific release can be identified.
- Not affected branches: a bulletin can list one current branch as not affected while older branches require patches. The overall label remains “Patch available”, but the correct action differs by branch.
Use the affected product name, affected range and matching fixed row together. If any one of those is missing, confirm the release in the linked vendor advisory.
Are uploaded CVE lists retained?
No. CSV and TXT files are read in the browser so the page can extract valid CVE identifiers. The original file is not uploaded. When you start the lookup, only the normalized CVE identifiers are sent for that request. BlackTree does not write the submitted list or generated comparison result to the catalogue database and does not retain it by default.
The result remains visible in the current browser page until you clear it, replace it or close the page. Copying a table places it on your device’s clipboard. Downloading CSV creates a file on your device. Those local copies are then governed by your browser, operating system, email application and organisation’s data-handling rules.
Use this function only for public CVE identifiers. Do not add hostnames, IP addresses, usernames, asset names, incident notes, credentials or other internal context to the uploaded file.
Why does a country tab contain source-language text?
Each country report separates BlackTree interface text from the wording supplied by the national authority. Navigation labels, field names and explanatory notes are presented in the report language. Advisory content is preserved as the source published it so that BlackTree does not silently alter a security statement.
A Dutch authority can publish a Dutch title and remediation note while quoting an English vendor description. In that case, the English passage is still part of the Dutch source and remains English. Similarly, product names, technical terms and vendor release titles may not have a meaningful local translation.
Use the displayed source-language label and official bulletin link to verify the original context. If a passage looks inconsistent, compare it with the attached source before treating it as a translation error. BlackTree should localise its own fallback phrases, but it should not invent a translation for authority-supplied evidence.
What does an unmatched SBOM component mean?
It means the matcher did not find a sufficiently reliable relationship between the component and a product in the current catalogue. It does not prove that the component is unaffected. Work through these checks before reaching a conclusion:
- Check exact identifiers. Compare the package URL (purl), CPE, SWID reference, package namespace and supplier name with the component’s official package or vendor page.
- Check common aliases. Look for renamed products, former vendor names, abbreviations, repository names and differences between a package name and its marketed product name.
- Check the ecosystem. Confirm that the package belongs to the expected ecosystem, such as npm, Maven, PyPI, NuGet, Debian or RPM. The same short name can refer to unrelated software in different ecosystems.
- Check the version. Remove build metadata only when the package rules permit it, and verify whether the vendor uses release branches, editions or platform-specific versions that differ from the SBOM value.
- Check authoritative evidence. Search the vendor advisory, release notes and the linked CVE record for the exact product family and affected version range.
If the component still cannot be matched, record it as unresolved and review it manually. Do not convert an unmatched result into “not affected”.
What does an ambiguous inventory match mean?
It means the matcher found a plausible relationship but not enough precise evidence to select one product or CVE automatically. This commonly happens when an SBOM contains only a short package name, when several vendors use the same name, when a project has been renamed or forked, or when the package version cannot be aligned confidently with the vendor’s release scheme.
Review every suggested candidate rather than choosing the first result. Compare the supplier, namespace, purl, CPE, repository, homepage, package ecosystem and version with the official component record. Then open the candidate CVE reports and check whether the affected product and version branch describe the component you actually have.
For example, a component called “gateway” could refer to unrelated packages in npm, Maven or a vendor appliance. A name match alone is too weak. A matching purl namespace, supplier and version range may resolve it; conflicting ecosystems or vendors should rule it out.
If the evidence remains incomplete, keep the component in manual review. An ambiguous match is not a confirmed vulnerability and is not an unaffected result.
