Run a scheduled job that pulls the NVD JSON feed on whatever cadence you choose (hourly, daily, weekly) and writes new and revised entries into your sheet. New CVEs appear as rows; revised CVEs update existing rows; the page-group cache invalidates affected slugs. The corpus stays current within your refresh window automatically.

Yes. Filter the NVD feed at ingest time by CPE vendor or product, keeping only the slice of CVEs relevant to your audience. A Kubernetes-focused site might ingest only CVEs touching CPE strings containing kubernetes, etcd, or container runtimes, producing a 5,000-page targeted corpus instead of 200,000 generic entries.

Add a cvss_v4_score and cvss_v4_vector pair of columns. The base template renders both v3 and v4 chips when v4 is available, falling back to v3 only for older CVEs. The transition rolls out naturally as NVD adds v4 scoring to new entries, with no per-page work needed on the publisher side.

Each CVE row carries a cwe_mapping field. A link-template selector renders the CWE link into the CWE reference page on your site. The reverse direction works through the example_cves arrays on your CWE pages. Both endpoints read from each other's slug fields, so the cross-link graph stays bidirectional automatically.

Yes. The kev_status field drives a CISA KEV banner via conditional Twig in the base template. A separate exploit_db_ids column links to exploit-db.com entries when known. A metasploit_module column flags Metasploit module availability. The severity panel surfaces these flags at the top of every affected page for fast triage.

No. The blog at /blog/ continues to publish under its own template; the CVE corpus lives under /cve/ via SleekRank. Blog posts that deep-dive into a major CVE link to the canonical /cve/{slug}/ page; the CVE page links back to the deep-dive in a Coverage block populated through an editorial_links array on the row.

Add a state column from the NVD feed (RESERVED, PUBLISHED, REJECTED). A conditional mapping excludes RESERVED and REJECTED states from public render unless you want a stub page for reserved IDs. State transitions flow through the next NVD pull automatically, so reserved IDs publish as soon as their full disclosure appears upstream.

SleekRank doesn't materialize all pages at once. Pages render on-demand from the data source with the cache layer holding compiled HTML. The 200,000 entries live in the sheet and the SleekRank items cache table; only requested URLs render. Search engines crawl the sitemap and cache is populated organically as the crawler visits each page.