VulnSea is live
VulnSea is live. It's CVE and 0day intelligence built to be read by humans and AI agents alike — newly released vulnerabilities across every platform, presented as a fast, sleek feed that's verbose on demand.
Severity reads as depth
We aggregate from the NVD and GitHub Advisories, then enrich every record with CISA KEV status, EPSS exploitation-probability scores, and CVSS severity. Our depth model fuses severity with real-world exploitation evidence into a single signal — the deeper the contact, the graver the threat — so you can triage at a glance instead of squinting at a base score.
Agents are first-class
This is the part we care about most. Every CVE is served as raw markdown, and the whole corpus is reachable through:
- a documented REST API and an OpenAPI 3.1 spec (
/openapi.json) - a machine-readable manifest at
/llms.txt - a native MCP endpoint (
/api/mcp) for tool-use models - cost-control knobs — sparse
fields, batchids, and cheap freshness probes — built for LLM pipelines
Point your agent loop straight at it. The About page walks through patterns for automated triage, continuous vulnerability management, and on-demand risk assessment.
Free to read, free to dive
Every CVE is free to read and needs no account. VulnSea is ad-supported — there are no paid tiers. Sign in for saved searches, watchlists, API keys (higher rate limit), and alerts (Slack, Discord, or webhook).
Stay in the loop
Follow along as we ship — and break down the vulnerabilities worth your attention:
- Bluesky: @vulnsea.com
- LinkedIn: VulnSea
Sound the depths → vulnsea.com