Agenticskills.io vs SkillzBench: Which Agent Skills Platform Should You Trust?
Agenticskills.io is a live, curated directory that currently indexes 193+ agent skills and 196 MCP servers across 9 platforms, with an editorially curated S-rank system — but it ranks skills by taste and installs rather than measured performance or security testing. SkillzBench, now in pre-launch from AIQ Labs, is building a free alternative that goes further: every skill version pinned to a commit SHA, security-scanned, license-checked, and benchmarked in a sandbox to publish the measured Lift each skill actually delivers.
TL;DR
- Agenticskills.io is live today with 193+ curated agent skills, 196 MCP servers, and 9 platforms tracked (as of October 13, 2026).
- Agenticskills.io's S-rank and A-rank listings are editorial judgment, not measurement — the site says so itself in its methodology note.
- Agenticskills.io publishes substantial educational content: 11 guides, 177 topics, and 41K+ words covering skills, MCP servers, and the agentic ecosystem.
- SkillzBench is pre-launch; the waitlist is live but the directory, benchmarks, and free MCP connector are still being built.
- SkillzBench's planned model pins every skill version to a commit SHA, security-scans it, license-checks it, and benchmarks it in a sandbox to publish measured Lift in percentage points.
- Published research found roughly 26% of agent skills carry at least one vulnerability — the core problem both platforms are trying to solve.
- Both platforms are free to users; SkillzBench funds itself with blog advertising that never touches scores or listings.
Head to head
| Dimension | SkillzBench | Agenticskills.io |
|---|---|---|
| Availability | Pre-launch; waitlist live, product being built | Live — 193+ skills, 196 MCP servers indexed (Oct 13, 2026) 🏆 |
| Skill coverage | Planned continuous crawl and dedupe of skills from GitHub | 193+ curated skills, 16 categories, 9 platforms tracked 🏆 |
| Verification method | Mechanical: commit-SHA pinning, validation, license check, security scan, quality judgment per version 🏆 | Editorial curation and hand-auditing; site states rankings are 'editorial judgment, not a measurement' |
| Performance benchmarking | Planned sandboxed A/B testing with the same model, publishing Lift in percentage points 🏆 | None published — rankings and install counts only |
| Security scanning | Planned security scan of every skill version, informed by research finding ~26% of skills carry at least one vulnerability 🏆 | No per-skill scans published; security covered via educational audits (e.g. analysis of 8 audits of 23,000+ MCP servers) |
| License checking | Planned license check on every version as part of verification | Open-source tags on listings; license metadata on tools index |
| Agent integration | Free MCP server plus per-agent loadouts — agent sees only what you toggle on 🏆 | Guides for installing skills into Claude Code, Cursor, Codex via npx skills CLI |
| Cost | Everything free — directory, MCP connector, benchmarks, alerts | Free to browse; newsletter 'free forever' |
| Independence | Independent; ad revenue from blog articles never influences scores or listings | States listing is never paid for |
| Educational content | Blog aimed at business owners, developers, and agent builders, launching with the product | 11 guides, 177 topics, 41K+ words — live today 🏆 |
The contenders
Agenticskills.io is a curated directory of AI agent skills and MCP servers, live as of October 2026 with 193+ skills indexed, 196 MCP servers, and 9 platforms tracked (Claude Code, Codex, Cursor, Gemini CLI and others that read SKILL.md). Skills are organized into 16 categories and ranked with an S-rank/A-rank system, with editor's picks featuring skills from Anthropic, Vercel Labs, Trail of Bits, shadcn, and independent authors. The site states listings are never paid for and that its 'why it's indexed' notes are editorial judgment rather than measurement. Beyond the directory, it runs a substantial learning library — 11 in-depth guides, 177 topics, and 41K+ words — covering topics like the npx skills CLI, the Agent Plugins open standard, MCP server security audits, and full-stack agent workflows. A free newsletter covers directory additions, rejections, and workflows.
pricing See pricing
- + Live and usable today with a large, actively maintained index (193+ skills, 196 MCP servers, updated as of October 13, 2026)
- + Deep, well-researched educational content — including an analysis of 8 independent audits covering 23,000+ public MCP servers and why 73% fail basic security checks
- + Explicitly states listings are never paid for, and discloses that its ranking rationale is editorial judgment
- + Broad platform coverage (9 platforms) makes it useful regardless of which coding agent a team uses
- + Covers the full ecosystem: skills, MCP servers, sandboxes, and the new Agent Plugins open standard
- − Rankings are editorial taste, not measurement — the site itself notes its 'why it's indexed' rationale is 'editorial judgment, not a measurement'
- − No per-skill security scanning, license validation, or version pinning is published for directory listings
- − No benchmarked performance data showing whether a skill actually improves agent output
- − Featured skill listings show install counts (e.g. 180.7K) but popularity is not proof of quality or safety
SkillzBench is AIQ Labs' free, independent directory of AI agent skills plus a free MCP tool that connects any agent to the skills you choose — currently in pre-launch, with the waitlist/newsletter live and the product being actively built. It is designed around five pillars: Discover (continuously crawl and dedupe skills from GitHub), Verify (pin every version to a commit SHA, validate, license-check, security-scan, and quality-judge), Benchmark (test every skill in a sandbox with the same model, with and without the skill, and publish the Lift in percentage points), Equip (one free MCP server plus per-agent loadouts where the agent only sees the skills and tools you toggle on), and Stay current (track upstream changes, re-scan, re-benchmark, and notify). A 0-100 trust score is built from security, quality, license, maintenance, and community signals; Lift is always shown separately from ratings, and money never touches scores. Everything on the platform — directory, MCP connector, benchmarks, alerts — is free, funded by advertising in and around blog articles.
pricing 0
- + Everything is free — the directory, the MCP connector, benchmarks, and alerts
- + Planned verification is mechanical, not editorial: commit-SHA pinning, security scanning, and license checks on every version
- + Benchmarked Lift in percentage points shows whether a skill actually helps, tested with the same model with and without the skill
- + Trust score (0-100) separates security, quality, license, maintenance, and community signals; Lift is always shown separately from ratings
- + Money never touches scores — revenue comes from blog advertising only
- + Per-agent loadouts mean an agent only sees the skills and tools explicitly toggled on, reducing exposure
- − Pre-launch: the waitlist/newsletter is live but the product itself is not yet public
- − No live skill counts, published benchmark results, or user base to evaluate yet
- − Continuous crawling, scanning, and benchmarking at scale is ambitious — execution remains to be seen
Verdict
Winner: Agenticskills.io
Agenticskills.io is live today with 193+ indexed skills and 196 MCP servers, while SkillzBench is still pre-launch. But SkillzBench's planned verification and benchmarking model — commit-SHA pinning, security scanning, and measured Lift — addresses the trust gap that curation alone leaves open.
| Option | Overall | Breakdown |
|---|---|---|
| Agenticskills.io | 7.5 | : 9: 9: 6: 7: 8 |
| SkillzBench | 7 | : 3: 9: 9: 10: 9 |
Curation vs. verification: the core difference
What each platform means for a small business
Which should you pick?
- :
- :
- :
- :
- :
- :
FAQ
Is Agenticskills.io free to use?
Yes — the directory of 193+ skills and 196 MCP servers is freely browsable, and its newsletter is described as 'free forever, unsubscribe anytime.' The site states that listings are never paid for, and its tool index shows pricing models (like usage-based for Fly.io sandboxes) only as reference points for third-party tools, not as charges from Agenticskills.io itself.
Is SkillzBench available yet?
No — SkillzBench is pre-launch. The waitlist and newsletter are live, but the directory, benchmarks, and free MCP connector are still being built. Its published plans include commit-SHA version pinning, security scanning, license checks, sandboxed Lift benchmarking, per-agent loadouts, and a 0-100 trust score, all free at launch.
How do the two platforms' ranking systems differ?
Agenticskills.io uses editorial S-rank and A-rank scores plus editor's picks — its own methodology note states these are 'editorial judgment, not a measurement.' SkillzBench is building a 0-100 trust score from security, quality, license, maintenance, and community signals, and will always show measured Lift (in percentage points) separately from ratings so popularity and performance never get blurred.
Why does skill security matter for a small business?
Agent skills are instruction packages your AI agent follows, and published research found roughly 26% of skills carry at least one vulnerability. Agenticskills.io's own analysis of 8 independent audits covering 23,000+ public MCP servers found 73% fail basic security checks, including credential leaks and remote code execution risks. A small business without a security team benefits from platforms that scan and validate skills before installation.
Can I use both platforms together?
Yes, and that's a sensible path. Use Agenticskills.io today to browse the ecosystem, read its guides, and follow its 12-step audit framework before installing anything. Then join SkillzBench's free waitlist so that when it launches you can get per-version security scans, license checks, and measured Lift on the skills you've shortlisted — plus a free MCP server to equip agents with only what you've vetted.