Question 1

How does SkillzBench security-scan AI agent skills for vulnerabilities?

Accepted Answer

SkillzBench's Verify pillar pins every skill version to a commit SHA, then validates it, runs license checks, performs security scanning and applies quality judging. A 0-100 trust score combines security, quality, license, maintenance and community signals so you can assess risk at a glance. Note that SkillzBench is pre-launch and actively being built — join the waitlist for access as these capabilities go live.

Question 2

Can SkillzBench stop a risky skill from reaching my marketing team's AI agents?

Accepted Answer

Yes, by design. The Equip pillar uses one MCP server with per-agent loadouts — you toggle skills and tools per agent, and the agent only sees what was chosen. Script-bearing skills are listed link-out only at launch, and instruction-only skills (SKILL.md plus text references) are fully verified, so nothing unvetted gets equipped through SkillzBench.

Question 3

Does SkillzBench help with audit and compliance requirements for AI tooling?

Accepted Answer

Every skill version is pinned to a commit SHA, giving you a reproducible record of exactly what your agents ran. Teams-tier plans (hypothesized at ~$15-25/seat/mo) are designed to include org loadouts, allowlists and audit capabilities. Combined with the 0-100 trust score, you get documented evidence of security, quality and license posture for each equipped skill.

Question 4

When will SkillzBench be available, and what does pricing look like?

Accepted Answer

SkillzBench is pre-launch — the waitlist/newsletter is live now, with the product being actively built. The pricing hypothesis is a generous free tier, Pro at roughly $9-15/mo for more agents and unlimited enabled skills, Teams at roughly $15-25/seat/mo for org loadouts, allowlists and audit, Creator at roughly $5-10/mo for verified authors, plus benchmark-as-a-service for priority runs.

Question 5

How is SkillzBench different from just letting developers pick skills from GitHub?

Accepted Answer

GitHub gives you a README and stars; SkillzBench gives you independent evidence. Every skill is deduped from continuous GitHub crawling, pinned to a commit SHA, security-scanned, license-checked, quality-judged, and benchmarked in a sandbox with the same model — with and without the skill — publishing Lift in percentage points. And because SkillzBench is an independent lab where money never touches scores, the numbers aren't influenced by who published the skill.

Agent Security Scanning for Marketers