Question 1

How does SkillzBench handle license checks on third-party AI agent skills?

Accepted Answer

SkillzBench's Verify pillar pins every skill version to a commit SHA, then validates, license-checks, security-scans and quality-judges it. License signals also feed into the 0-100 trust score alongside security, quality, maintenance and community signals, so product managers can see licensing risk at a glance before adoption. SkillzBench is pre-launch; the waitlist is live now.

Question 2

Can we prove to auditors that the AI skills our teams use are secure and compliant?

Accepted Answer

Yes — that's the goal of the Teams tier, which is being built with org loadouts, allowlists and audit capabilities. Combined with commit-SHA pinning, security scans and a documented trust score, you get a verifiable record of what was reviewed, when, and against which version. Note the product is not yet publicly launched.

Question 3

What happens when a skill we've approved changes upstream on GitHub?

Accepted Answer

The Stay-current pillar tracks upstream changes, then re-scans and re-benchmarks the skill and notifies you. Because every version is pinned to a commit SHA, an upstream change can't silently alter what your agents run — you'll always know a new version needs re-review.

Question 4

Does SkillzBench execute skill scripts, and what about script-bearing skills?

Accepted Answer

Launch scope is instruction-only skills — SKILL.md packages plus text references. Script-bearing skills are listed link-out only, meaning SkillzBench surfaces them with metadata but does not execute their code, keeping the verification surface controlled and safe.

Question 5

How is SkillzBench different from just reading a skill's GitHub README and star count?

Accepted Answer

GitHub stars and READMEs are self-reported; SkillzBench is an independent lab. It benchmarks every skill in a sandbox with the same model, with and without the skill, publishes Lift in percentage points, and computes a 0-100 trust score from security, quality, license, maintenance and community signals. Money never touches scores, and Lift is always shown separately from ratings.

Skill Compliance & License Checks for Product Managers