Question 1

Does SkillzBench work with Claude Code, Codex, Cursor, and Gemini CLI agents?

Accepted Answer

Yes — SkillzBench is being built around the Agent Skills standard (SKILL.md instruction packages) for Claude Code, Codex, Cursor, Gemini CLI, and chat-app agents. Equip delivers one MCP server plus per-agent loadouts, and each agent only sees the skills and tools you've toggled on for it. Note that launch scope covers instruction-only skills (SKILL.md plus text references); script-bearing skills are listed link-out only.

Question 2

How does SkillzBench measure whether a skill actually improves an agent?

Accepted Answer

Every skill is benchmarked in a sandbox using the same model, tested both with and without the skill, and SkillzBench publishes the resulting Lift in percentage points. Lift is always shown separately from ratings and Trust scores, and money never touches scores — so the number you see is a measurement, not a marketing claim.

Question 3

What does the Trust score actually take into account?

Accepted Answer

Each skill gets a 0-100 Trust score built from security, quality, license, maintenance, and community signals. Every version is pinned to a commit SHA and put through validation, license checks, security scans, and quality judging. This matters because published research found roughly of skills carry at least one vulnerability — the score exists so agencies don't have to gamble on client-facing agents.

Question 4

Is SkillzBench available now, and how do agencies get access?

Accepted Answer

SkillzBench is pre-launch and actively being built — the waitlist and newsletter are live, but the product is not yet public. Join the waitlist to be notified as the Discover, Verify, Benchmark, Equip, and Stay-current pillars come online. Pricing under consideration includes a generous free tier, a Pro tier (~$9-15/mo) with more agents and unlimited enabled skills, and a Teams tier (~$15-25/seat/mo) with org loadouts, allowlists, and audit.

Question 5

How is this different from just installing skills from GitHub or a marketplace directly?

Accepted Answer

Raw GitHub installs give you no version pinning, no security scan, no license check, and no evidence the skill helps. SkillzBench continuously crawls and dedupes skills from GitHub, pins every version to a commit SHA, validates and scans it, benchmarks it with and without the skill to publish Lift, and then lets you equip agents through controlled per-agent loadouts via one MCP server. Marketplaces rank by popularity; SkillzBench measures performance and safety — and money never touches scores.

Curated Agent Loadouts for Agencies