Feeds & discovery
vulnetix_soc_triage
The score-driven queue: recently published vulnerabilities ranked for attention, each with CVSS, EPSS and CESS scores plus vendor and product.
| Argument | Type | Default |
|---|---|---|
limit | 1–100 | 25 |
offset | number | 0 |
What is worth looking at today? Cross-reference the top 10 against this repo's dependencies.
vulnetix_search
Full-text search across titles, descriptions and aliases, ranked, with the excerpt that matched.
| Argument | Type | Default |
|---|---|---|
query | string, required | none |
limit | 1–50 | 10 |
offset | number | 0 |
For when you remember the nickname but not the identifier: “log4shell”, “spring4shell”, “that OpenSSL heap thing”.
vulnetix_eol_products
Products with tracked end-of-life dates (runtimes, servers, databases, operating systems) with release and maintained-release counts.
| Argument | Type | Default |
|---|---|---|
category | string | all |
limit | 1–500 | 150 |
Follow up with vdb_request on /eol/products/{product} for per-release dates.
An unsupported runtime is a vulnerability that will never be fixed, which is why this sits alongside the advisory tools rather than in a separate product.
vulnetix_threat_feed
The one composing tool. Five independent upstream calls, issued concurrently.
| Argument | Type | Default |
|---|---|---|
limit | 1–50 | 10 |
sections | array of section names | all five |
Sections: ai-discoveries, ai-in-wild, ai-malware, exploit-trends,
vendor-trends.
Partial failure degrades, it does not fail. A leg that errors reports itself
in _meta["vulnetix/degraded"] and the other four still return. A five-way
digest that fails whole because one upstream was slow would be useless.
Give me a threat digest for this week's security newsletter.
Restrict when you only want part of it:
{ "sections": ["exploit-trends", "vendor-trends"], "limit": 20 }
What each section carries
| Section | Contents |
|---|---|
ai-discoveries | Vulnerabilities found by AI systems, with the org, harness and whether a human assisted |
ai-in-wild | AI-assisted exploitation observed in the wild |
ai-malware | AI-generated malware families, grouped by use mode |
exploit-trends | Totals by weaponisation tier, plus top CVEs per tier |
vendor-trends | CVE volume by vendor, current period leaderboard with prior-year totals for movement |
Trend series are truncated to recent periods; the note in the response says so
and points at vdb_request for the full history.