CVE & exposure intelligence, inside your AI tools

Counts of active CVEs by severity band, the count with no severity band from any source (not yet scored, not a rating of None), and when the feed was last updated.

Ask: “How many critical CVEs does the feed hold, and when was it last updated?”

Search/filter CVEs (severity, min CVSS, text, sort) with EchelonGraph scores. A CVE not yet scored comes back NOT YET SCORED (score_assessed false), not 0.

Ask: “Find recent CVEs affecting nginx, with EchelonGraph scores.”

Full detail for one CVE — description, CVSS v3/v4, EG score, EPSS, KEV + ransomware, GHSA, references. A CVE not yet scored comes back NOT YET SCORED (score_assessed false), not 0.

Ask: “Give me the full picture on CVE-2024-3094.”

Internet-exposure footprint for a CVE: how many internet-facing services (distinct ip:port) EchelonGraph's KEV-exposure radar has on record running a version that maps to it, with a country/product breakdown. Derived from Shodan data. Shodan data is owned by Shodan, which holds its copyright (© Shodan). EchelonGraph claims no ownership of it or copyright in it. A CVE outside the radar's tracked set comes back NOT ASSESSED (exposure_state not_assessed), not 0.

Ask: “How many exposed services does EchelonGraph's radar have on record for CVE-2023-44487?”

Aggregate totals across the exposure radars: shadow AI; services running CISA-KEV CVEs; unauthenticated data stores and observability UIs, found through Shodan (LeakIX when Shodan query credits run low) and each confirmed by EchelonGraph's own identified check; leaked credentials; and MCP-server counts: hostnames named like an MCP server in EchelonGraph's own Certificate Transparency feed (no Shodan data), each counted once by its latest verdict from EchelonGraph's identified MCP probe, which never sends tools/call. Shodan data is owned by Shodan, which holds its copyright (© Shodan). EchelonGraph claims no ownership of it or copyright in it.

Ask: “How many unauthenticated data stores has EchelonGraph's radar confirmed?”

The CVEs CISA has added to its Known Exploited Vulnerabilities (KEV) catalog, newest first, with kev_added_date, kev_due_date and kev_ransomware (known ransomware-campaign use), from EchelonGraph's copy of that catalog, which polls CISA's feed every 5 minutes. Filter by since, until, ransomware and vendor; page with limit and cursor.

Ask: “Which CVEs has CISA added to its KEV catalog newest first, with known ransomware-campaign use?”

How one CVE's EPSS score (FIRST's exploit-prediction probability, with its percentile) has changed, as EchelonGraph recorded it: points, oldest first, and current, the record's value now. The series is change-only: a point is a recorded change, never a daily series, and before series_starts_at nothing was recorded.

Ask: “How has CVE-2023-44487's EPSS score changed, as EchelonGraph recorded it?”

Whether a product or package at a given version is affected by known CVEs, from the same matcher as echelongraph.io/am-i-affected. The CPE path takes product and version; the registry path takes ecosystem (npm, PyPI, Maven and other OSV ecosystem names), package and version. count depends on assessed: assessed false means the lookup did not evaluate this component, and a count of 0 there is not a finding of not affected. A registry match also carries interval, the advisory interval holding it, and fixed_in, the lowest unaffected version above it, or null with fixed_in_reason; a CPE match carries no fixed_in.

Ask: “Is nginx at a given version affected by known CVEs?”

Check a dependency list (purls, or a CycloneDX JSON or SPDX JSON document) against EchelonGraph's advisory corpus, one verdict per component: affected, not_affected, undetermined or not_assessed. Each row the lookup ran for carries matched_package and matched_via (a deb, apk or alpine purl's upstream qualifier matches as its source package; a refused row has neither), and each match carries fixed_in (the next unaffected version) or null with fixed_in_reason. For container images and Kubernetes pods, the input is an SBOM of each image, or its purls. Only the purls are sent to the API, never in a URL. Run from npm, this server is on your machine; over the hosted endpoint (mcp.echelongraph.io) it is EchelonGraph's, and the document is the request body. A deb, apk or rpm purl without a distro qualifier naming its release is not assessed, and not_assessed is never clean.

Ask: “Check this CycloneDX SBOM against the advisory corpus.”

Check a project's lockfile or pinned manifest against EchelonGraph's advisory corpus, one verdict per dependency: affected, not_affected, undetermined or not_assessed. Read: requirements.txt (== and === pins only), go.mod, package-lock.json, Cargo.lock, Gemfile.lock, composer.lock, poetry.lock and gradle.lockfile; go.sum is refused. Only the purls are sent to the API, never in a URL. Run from npm, this server is on your machine; over the hosted endpoint (mcp.echelongraph.io) it is EchelonGraph's, and the files are the request body. not_checked entries are not clean.

Ask: “Check this package-lock.json against the advisory corpus.”

Weakness, exploit code, detection checks, affected packages and fixed versions for one CVE, from EchelonGraph's per-CVE enrichment: cwes, exploits (at most 10, verified first) with exploits_total and exploits_by_kind, affected_packages and fixed_versions. Each affected package carries fixed_branches, every affected range with its fix: fixed_version is one range's fix, not the fix for every affected range. An empty exploits list is not evidence that no public exploit exists.

Ask: “Exploit code, detection checks, affected packages and fixed versions for CVE-2021-44228?”

How one CVE is fixed, as its sources state it, in one answer: cisa (for a KEV-listed CVE, CISA's required_action, verbatim; CISA's required_action is directed at US federal civilian agencies (under BOD 26-04, which replaced BOD 22-01 on 10 June 2026) and is not EchelonGraph's advice), fixed_branches (each affected package with every affected range), vendor_remediations (each vendor advisory naming the CVE, with remediation_state and items, each with kind, the vendor's own category, never read from the text), fix_references and patches. Every text is relayed as stated by its source; EchelonGraph has not tested it. An empty list or a remediation_state of none_in_source or not_parsed is not a finding that no fix exists.

Ask: “How is CVE-2021-44228 fixed, as its sources state it?”

One CWE (weakness class) and the CVEs classified under it: name and description from the MITRE CWE catalog, total, and one page of 50 CVEs, CISA-KEV-listed first, then EchelonGraph score. A total of 0 says that no CVE in EchelonGraph's feed is classified under that CWE, not that none exists.

Ask: “The CVEs classified under CWE-79, CISA-KEV-listed first.”

The vendor-published advisories that name one CVE, newest first, at most 20. Covers the vendor feeds EchelonGraph polls, for example Microsoft MSRC, Red Hat, Cisco, Palo Alto Networks and GitHub GHSA. Each advisory carries vendor_published_at (the vendor's date), our_first_seen_at (when EchelonGraph first recorded it) and withdrawn (the vendor rescinded it).

Ask: “The vendor advisories that name CVE-2024-21412, newest first.”

One vendor advisory in full, by vendor and the vendor's advisory ID: title, description, severity, cve_ids, affected_products, remediation and references, and withdrawn_at and withdrawn_reason when the vendor withdrew it.

Ask: “The remediation and references of one vendor advisory, by vendor and advisory ID.”

Search or list vendor-published advisories, newest first: a query of 3 or more characters is matched case-insensitively as a substring of each advisory's title, description, vendor name, vendor_advisory_id, affected_products and cve_ids, and a query of 1 or 2 characters matches whole words only; filter by vendor, by severity band, and by whether the advisory names any CVE ID. Advisories their vendor withdrew are left out. The query is sent in a request header, never in the URL.

Ask: “Search vendor advisories by severity band, newest first.”