Question 1

How does SkillzBench verify a SKILL.md skill version?

Accepted Answer

SkillzBench is building a Verify pillar that pins every version to a commit SHA, then validates the skill, license-checks it, security-scans it, and quality-judges it. That means the exact version your agency ships is the exact version that was scanned — no bait-and-switch between install and update.

Question 2

What does the benchmark actually measure for agent skills?

Accepted Answer

Each skill is tested in a sandbox with the same model, with and without the skill, and SkillzBench publishes the Lift in percentage points. This is designed to answer the question agencies actually care about: does this skill measurably improve agent performance, or is it just another markdown prompt bloating context?

Question 3

Does SkillzBench handle script-bearing skills, or only instruction-only SKILL.md packages?

Accepted Answer

Launch scope is instruction-only skills — SKILL.md plus text references. Script-bearing skills will be listed link-out only. If your agency builds skills with scripts, you can still surface them for discovery, but verification and benchmarking at launch focus on instruction-only packages.

Question 4

Is SkillzBench available now, and when can our agency start verifying skills?

Accepted Answer

SkillzBench is pre-launch and actively being built — the waitlist/newsletter is live, but the product is not yet public. Join the waitlist to be notified when verification, benchmarking, and the MCP server go live. Benchmark-as-a-service for priority runs is also planned.

Question 5

How is this different from just using a community marketplace's trust signals?

Accepted Answer

Marketplace install counts can be trivially spoofed via unauthenticated GET requests, and skills are often only re-scanned when popular. SkillzBench is an independent lab: a trust score 0-100 from security, quality, license, maintenance and community signals, Lift always shown separately from ratings, and money never touching scores. It also tracks upstream changes and re-scans and re-benchmarks when a skill updates.

Skill Author Verification for Agencies