Question 1

Does SkillzBench handle security scanning and license checks for AI agent skills used in security operations?

Accepted Answer

Yes — that's the core of the Verify pillar SkillzBench is building. Every skill gets pinned to a commit SHA, then license-checked, security-scanned, and quality-judged. Published research found roughly of skills carry at least one vulnerability, which is why verification happens before anything is enabled for an agent.

Question 2

How does SkillzBench measure whether an AI skill actually improves agent performance?

Accepted Answer

The Benchmark pillar tests every skill in a sandbox using the same model, running the task with and without the skill, and publishes the Lift in percentage points. Lift is always shown separately from trust ratings, so performance data and quality signals never get blended together.

Question 3

Can we control which AI skills each of our agents can access?

Accepted Answer

Yes. The Equip pillar provides one MCP server plus per-agent loadouts — you toggle skills and tools per agent, and the agent only sees what was chosen. This is especially useful for teams that want tighter control over what automation touches client-facing operations.

Question 4

Is SkillzBench available now, and what's the timeline for getting started?

Accepted Answer

SkillzBench is pre-launch and actively being built — the waitlist is live now and early access will roll out as verification and benchmarking come online. Launch scope covers instruction-only skills (SKILL.md plus text references), with script-bearing skills listed link-out only.

Question 5

How is SkillzBench different from just picking skills off GitHub or a marketplace?

Accepted Answer

SkillzBench is an independent testing lab, not a marketplace. It continuously crawls and dedupes skills from GitHub, pins versions to commit SHAs, runs license checks and security scans, benchmarks Lift with the same model, and tracks upstream changes with re-scans and notifications. Trust scores come from security, quality, license, maintenance, and community signals — and money never touches scores.

Skill Compliance & License Checks for Researchers