Question 1

Which agent platforms do SKILL.md skills from SkillzBench work with?

Accepted Answer

SkillzBench focuses on SKILL.md instruction packages built for Claude Code, Codex, Cursor, Gemini CLI and chat-app agents. At launch, the scope is instruction-only skills (SKILL.md plus text references); script-bearing skills are listed link-out only so you can evaluate them directly at the source.

Question 2

How does SkillzBench calculate the Lift metric for a skill?

Accepted Answer

Every skill is tested in a sandbox with the same model, once with the skill and once without. The difference in performance is published as the Lift in percentage points. Lift is always shown separately from ratings, and money never touches scores — so the number reflects measured performance, not sponsorship.

Question 3

Can SkillzBench help with AI agent security and compliance concerns my clients raise?

Accepted Answer

SkillzBench addresses the skill layer specifically: security scans, license checks and validation run on every verified skill, and trust scores combine security, quality, license, maintenance and community signals. Per-agent loadouts enforce least-privilege access — each agent only sees the skills and tools you toggled on. It complements, not replaces, your broader client security program.

Question 4

When can I actually start using SkillzBench?

Accepted Answer

SkillzBench is pre-launch and actively being built — the waitlist and newsletter are live now, but the product is not yet public. Joining the waitlist gets you notified as each of the five pillars (Discover, Verify, Benchmark, Equip, Stay current) rolls out, with early access for waitlist members.

Question 5

Why not just review GitHub skills manually or use the vendor's own docs?

Accepted Answer

Manual review doesn't scale past a handful of skills, and vendor docs are marketing, not evidence. SkillzBench is an independent lab: it dedupes the GitHub landscape, pins versions to commit SHAs, runs security and license checks, and publishes sandbox benchmark Lift scores. You get reproducible evidence instead of a README promise.

Agent Skill Migration for Consultants