What problem does it solve? CVSS scores measure theoretical severity, so digests sorted by score bury the vulnerabilities that are actually being exploited. This Skill inverts the ranking: it leads with confirmed exploitation from CISA KEV, enriches with EPSS exploitation probability, filters GitHub Advisories to your tracked stack, and outputs one concrete action per item. ## Core Features & Use Cases - Exploitation-first ranking: Pulls CISA KEV additions from the last 7 days and FIRST.org EPSS scores, then sorts items into PATCH TODAY, PATCH THIS WEEK, and MONITOR tiers. - Stack-aware filtering: Filters GitHub Advisory Database entries (last 48h, critical/high) to your tracked ecosystems (npm, pip, Go, crates.io, GitHub Actions), while KEV-listed CVEs always pass through. - Deduplication and logging: Skips CVE/GHSA IDs already reported in the last 2 days of memory logs, sends a formatted digest via ./notify, and logs tier counts and source status. - Use Case: A developer runs the digest each morning and gets a verdict line like "1 actively exploited, 2 likely soon" with exact upgrade commands such as "upgrade django to >=5.2.4" instead of a raw list of critical CVEs. ## Quick Start Run the security digest for my tracked ecosystems npm and pip and tell me what to patch today.
