skills/mukul975/anthropic-cybersecurity-skills/performing-cve-prioritization-with-kev-catalog/Gen Agent Trust Hub
performing-cve-prioritization-with-kev-catalog
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads the Known Exploited Vulnerabilities (KEV) catalog and EPSS scores from official CISA and FIRST.org sources to enrich vulnerability data.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted vulnerability identifiers and scan metrics from external files to perform its risk calculations.\n
- Ingestion points: Vulnerability data is read from user-provided files or strings in
scripts/agent.pyandscripts/process.py.\n - Boundary markers: Absent; the skill lacks specific delimiters to isolate external data within the logic flow.\n
- Capability inventory: The skill performs network GET operations using
requests.getand local file writing usingjson.dumpandpandas.to_csvinscripts/agent.pyandscripts/process.py.\n - Sanitization: External data is normalized to uppercase strings and cast to floating-point numbers in
SKILL.mdandscripts/process.pybefore being used in calculations.
Audit Metadata