Question 1

How does SkillzBench handle script-bearing skills that could execute code in our environment?

Accepted Answer

Launch scope is instruction-only skills — SKILL.md packages plus text references. Script-bearing skills are listed link-out only, so nothing executable is served through the MCP server. This keeps the attack surface of governed loadouts limited to validated instructions.

Question 2

Can we restrict which skills each analyst's agent can use across client engagements?

Accepted Answer

Yes. The Equip pillar provides per-agent loadouts: you toggle skills and tools per agent, and the agent only sees what was chosen. The Teams tier (~$15-25/seat/mo) adds org loadouts, allowlists and audit trails for agency-wide governance.

Question 3

How do you keep benchmark results trustworthy when skill authors might pay for placement?

Accepted Answer

Lift is always shown separately from trust scores, and money never touches scores. Benchmarks run in a sandbox with the same model, with and without the skill, and the Lift is published in percentage points. The trust score (0-100) is computed from security, quality, license, maintenance and community signals — independent of commercial interest.

Question 4

What's the timeline for agencies that join now, given the product is pre-launch?

Accepted Answer

SkillzBench is actively being built; the waitlist and newsletter are live today. Early access rolls out as the five pillars — Discover, Verify, Benchmark, Equip and Stay current — come online, with a generous free tier at launch and Teams pricing around $15-25/seat/mo for agency governance features.

Question 5

How is this different from just reading a skill's GitHub README and stars before installing it?

Accepted Answer

GitHub stars measure popularity, not safety or performance. SkillzBench pins every version to a commit SHA, runs license checks, security scans and quality judging, then benchmarks the skill in a sandbox to measure actual Lift. It also continuously tracks upstream changes and re-scans and re-benchmarks when a skill you depend on is updated — something no manual review process sustains.

Team Skill Governance for Agencies