Question 1

How does SkillzBench measure whether an agent skill actually improves my agent's performance?

Accepted Answer

SkillzBench benchmarks every skill in a sandbox using the same model, tested with and without the skill, and publishes the resulting Lift in percentage points. Lift is always shown separately from trust ratings, so you can see raw performance impact independent of quality or security scoring.

Question 2

What's the difference between agent skills and MCP servers, and where does SkillzBench fit?

Accepted Answer

Skills are SKILL.md instruction packages that give agents procedural knowledge via progressive disclosure, while MCP servers expose tools. SkillzBench covers both worlds: it verifies and benchmarks instruction-only skills (script-bearing skills are listed link-out only), and it equips everything through one MCP server with per-agent loadouts where you toggle skills and tools per agent.

Question 3

Are script-bearing skills supported for benchmarking?

Accepted Answer

At launch, SkillzBench's scope is instruction-only skills — SKILL.md files plus text references. Script-bearing skills are listed link-out only, so you can discover them but SkillzBench won't benchmark or scan them initially.

Question 4

How do I know a skill won't break or go stale after I add it to my loadout?

Accepted Answer

SkillzBench's Stay Current pillar tracks upstream changes, re-scans and re-benchmarks skills, and notifies you. Every version is pinned to a commit SHA, so you always know exactly which version you're running and whether it has changed upstream.

Question 5

Can I trust the ratings, or do skill authors pay for better placement?

Accepted Answer

Money never touches scores. Lift is always shown separately from ratings, and the trust score (0-100) is built from security, quality, license, maintenance and community signals. SkillzBench is an independent testing lab — AIQ Labs' dedicated benchmarking arm — not a skill marketplace with sponsored rankings.

Curated Agent Loadouts for Freelancers