Question 1

How does SkillzBench tell me if an agent skill from GitHub is safe to use?

Accepted Answer

SkillzBench is building a Verify pillar that pins every skill version to a commit SHA, then validates it, checks its license, security-scans it and quality-judges it. Results feed a 0-100 Trust Score built from security, quality, license, maintenance and community signals. Note that launch scope covers instruction-only skills (SKILL.md plus text references); script-bearing skills are listed link-out only.

Question 2

What does 'Lift' actually mean when benchmarking a SKILL.md package?

Accepted Answer

SkillzBench's Benchmark pillar tests every skill in a sandbox using the same model, running the task with and without the skill installed. The difference is published as the Lift in percentage points. Lift is always shown separately from ratings and scores, and money never touches either — so you see genuine performance impact, not paid placement.

Question 3

Can I manage different skill sets for Claude Code versus Cursor versus Gemini CLI?

Accepted Answer

Yes — the Equip pillar is designed around one MCP server plus per-agent loadouts. You toggle skills and tools per agent, and each agent only sees what you chose for it. That means your Claude Code setup can differ from your Codex or Gemini CLI setup without cross-contamination of instructions or tools.

Question 4

Is SkillzBench available now, and what will it cost a solo freelancer?

Accepted Answer

SkillzBench is pre-launch and actively being built — the waitlist is live, the product is not yet public. The pricing hypothesis includes a generous free tier, with Pro at roughly $9-15/mo for more agents and unlimited enabled skills, and a Creator tier at roughly $5-10/mo for verified skill authors.

Question 5

Why not just read the README and star counts before installing a skill?

Accepted Answer

Star counts and READMEs tell you nothing about security — published research found roughly of skills carry at least one vulnerability. SkillzBench replaces vibes with commit-SHA-pinned verification, license checks, security scans, a 0-100 Trust Score, and measured Lift from sandbox benchmarks with the same model. Upstream changes are tracked too, so a skill that drifts or breaks gets re-scanned and re-benchmarked with a notification.

Agent Skill Migration for Freelancers