Agent Skils Hub is live · free · last verified 30 Oct 2026 · viability monitor
Agent Skils Hub
Search 255,000+ AI agent skills and MCP servers, each security-graded before you install it.
If you install MCP servers or Claude skills written by strangers, a security grade plus a quality score before the code touches your agent is the cheapest due diligence there is — and the whole thing is MIT-licensed. Treat the scan as a first filter, not an audit: UNSAFE is a hard stop, SAFE means read the diff yourself. The site's own census makes the case better than any pitch page: 1 in 32 graded skills came back unsafe.
Verdict generatedVerified 4h ago · liveness 66/100 · cite: rightaichoice.com/tools/agent-skils-hub
- Developers vetting an MCP server or Claude skill before running it in an agent
- Security-conscious teams needing a fast risk filter across hundreds of open-source agent tools
- Solo builders comparing browser, database or scraping tools for an AI coding stack
- Researchers citing agent-skill security and quality data from an open CC-BY-4.0 dataset
- Teams that need a formal security audit or a vendor-backed guarantee
- Buyers who need enterprise contracts, SLAs or a dedicated support contact
- Users who want hands-on setup help for individual tools
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Agent Skills Hub if you need a signed audit report, vendor-backed SLA or someone to support the tool after you adopt it — this is a rule-based scan over GitHub repos, run by one person as an open-source project.
A SAFE grade is a filter, not a clearance — you still need to read the code of anything you give credentials to, which is real time you budget for separately.
The whole directory is free and MIT-licensed, which puts it below paid alternatives like commercial security scanning vendors or vendor-curated marketplaces. The trade-off isn't subscription cost, it's depth: it's a first-layer scan with no support contract, so teams that need an audited assessment pay for that elsewhere.
In short
Agent Skils Hub — Search 255,000+ AI agent skills and MCP servers, each security-graded before you install it. Best for Developers vetting an MCP server or Claude skill before running it in an agent, Security-conscious teams needing a fast risk filter across hundreds of open-source agent tools, Solo builders comparing browser, database or scraping tools for an AI coding stack. Free to use.
What is Agent Skils Hub?
Agent Skils Hub — Search 255,000+ AI agent skills and MCP servers, each security-graded before you install it. Best for Developers vetting an MCP server or Claude skill before running it in an agent, Security-conscious teams needing a fast risk filter across hundreds of open-source agent tools, Solo builders comparing browser, database or scraping tools for an AI coding stack. Free to use.
Agent Skills Hub is an open-source directory of 255,000+ AI agent skills, MCP servers, Claude skills, Codex skills, prompt libraries and agent frameworks, rebuilt from GitHub every 8 hours. It targets the question most tool listings skip: is this thing safe to run inside your agent? Every indexed repository with a fetched README carries a grade of SAFE, CAUTION, UNSAFE or UNAUDITED from a rule-based scan on the SlowMist agent-security taxonomy, now spanning 42 published rules across credential harvesting, data exfiltration and curl | sh installers. Independent research cited on the site backs the concern: Liu et al. (2026), arXiv:2601.10338, put 26.1% of 31,132 agent skills as carrying security vulnerabilities, and the site's own census found 1 in 32 graded skills unsafe. A separate 0–100 quality score spans six dimensions — completeness, clarity, specificity, examples, README structure and agent readiness — so you can tell a finished project from a weekend experiment without reading hundreds of READMEs. Coverage is specific: 93% of skills over 100 stars and 98% over 1,000 are graded, and the census found risk concentrated in the long tail. Discovery runs as a 6-phase pipeline every 8 hours: GitHub search, metadata enrichment, README fetch, quality scoring, category classification and composability analysis, with no manual curation in the rankings. Browse seven categories, filter by language, hit scenario pages (browser automation, code review, database MCP servers, PPT skills, image generation), run tools head to head in Compare, or paste a repo into the Skill Analyzer for a grade plus platform compatibility checks. It's built for developers and security-minded teams vetting open-source tools, not for buyers who need a formal audit. The closest alternatives are plain GitHub search and marketplace listings, which rank by stars and installs. This ranks by whether the code is safe and actually finished.
Key features
- Security grades on indexed repos: SAFE, CAUTION, UNSAFE or UNAUDITED
- Rule-based scan built on the SlowMist agent-security taxonomy
- 42 published detection rules covering credential harvesting and data exfiltration
- curl | sh installer detection among the red-flag rules
- 0–100 quality score across six dimensions including agent readiness
- Index of 255,000+ GitHub repositories refreshed every 8 hours
- Categories: MCP Server, Claude Skill, Codex Skill, Agent Tool, Prompt Library, AI Coding Assistant, AI Tool
- Scenario pages with ranked shortlists — browser automation, code review, database MCP servers, PPT, image generation
- Compare view for evaluating multiple agent tools side by side
- Skill Analyzer returning security grades plus platform compatibility checks
- Filter by programming language and keyword search
- Weekly newsletter with top 20 movers by star velocity
- 6-phase pipeline: GitHub search, metadata enrichment, README fetch, quality scoring, classification, composability
- Full catalog published as a citable dataset (DOI 10.5281/zenodo.21292799, CC-BY-4.0) mirrored on Hugging Face
- Open-source code under MIT on GitHub
Behind the verdict
We'd reach for Agent Skills Hub the moment you're about to drop a stranger's MCP server into your agent config. The short version of what it buys you: a SAFE/CAUTION/UNSAFE grade and a 0–100 quality score in front of the install, instead of a star count that tells you nothing about whether the README's install line is a curl | sh pipe to somewhere unpleasant. For a free, MIT-licensed directory, that's a lot of sanity check for zero dollars. The honest comparison is GitHub search. GitHub ranks by stars and recency; it will happily surface a popular repo whose install script is sketchy or whose project died two commits in. AgentSkillsHub re-ranks on safety and completeness. The tradeoff is coverage in the tail — the census says only about 17.7% of indexed skills are popular enough to grade, so a long-tail repo you paste in may come back thin or UNAUDITED rather than fully scored. Where it bites: this is a rule-based first-layer scan, not a line-by-line human audit, and the methodology page says so. A SAFE grade means no red flags tripped, not that the code is trustworthy. Read the diff on anything you're actually deploying. And if you need a vendor-backed guarantee or an SLA, this isn't that — it's an independent solo-run project. Worth knowing before you lean on it long-term: the maintainer has been open about the economics. Strong traffic, $0 revenue, and a stated pivot from directory toward a trust layer for agent deployment (reported mid-2026). That doesn't change today's tool, but it's the kind of thing to keep an eye on if you're building a workflow around it. Pick it when you want a fast risk filter across hundreds of open-source agent tools, or when you need a citable dataset (DOI 10.5281/zenodo.21292799, CC-BY-4.0) for research. Pass on it if you're shopping
Limitations
- Security grading is a first-layer rule-based scan built on the SlowMist agent-security taxonomy, not a line-by-line human audit — a SAFE grade means no red flag fired, not that a person reviewed the code.
- Coverage is uneven: the June 2026 census reported 117,854 skills graded, only 17.7% popular enough to grade, and 1 in 32 graded skills unsafe, with risk concentrated in the long tail.
- Indexing covers GitHub open-source repositories only and refreshes every 8 hours rather than in real time, and ranking and scenario pages carry no manual curation.
- It is a solo, MIT-licensed project — in May 2026 the maintainer reported strong traffic with $0 revenue and announced a pivot from directory to trust layer, so the product's shape may move.
as of 2026-10-17
Researching Agent Skils Hub? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Is Agent Skils Hub still active?
Agent Skils Hub is live · free · last verified 30 Oct 2026 · viability monitor.
Viability score
How well maintained and how widely used is Agent Skils Hub? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →What's new in Agent Skils Hub
Checked 7 days agoAcross the latest 3 updates: 3 news mentions.
We Security-Graded 117,854 AI Agent Skills. Here's What We Found.
The full-catalog census reported 117,854 skills graded, only 17.7% popular enough to grade, and 1 in 32 graded skills unsafe, with risk concentrated in the long tail.
I Built an 86K AI Skill Directory in 10 Weeks. Here's Why I'm Changing Course.
The maintainer reported strong traffic but $0 revenue and announced a pivot from directory to a trust layer for AI agent deployment.
How I'd Build a SaaS Solo in 6 Weeks Using AI Skills (2026 Stack)
A guide laying out a five-tier AI skill stack for solo SaaS builds in six weeks, drawing examples from 19 hand-picked projects.
What do users say about Agent Skils Hub?
What people actually say about Agent Skils Hub — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
Researched26 mentions across 1 source (YouTube) · researched Sep 1, 2026.
Average across the 1 source that answered — each source counts once, not each post.
- +Security grading (SAFE/CAUTION/UNSAFE) gives instant trust signals before installing.
- +Huge catalog of 175,000+ tools across MCP, Claude skills, Codex, and more.
- +Open-source (MIT) and offers citable datasets on Zenodo/Hugging Face.
- +Automated refresh every 8 hours keeps listings current.
- +Quality scoring 0-100 across 10 dimensions helps compare tools.
- −Rule-based grading may overlook nuanced security risks.
- −Limited direct user feedback; most buzz is from related content.
- −No support for non-GitHub repositories, limiting coverage.
- −Some users may find the grading overly cautious or not cautious enough.
- −No clear community forum or direct support channel.
- • No explicit hidden costs mentioned in data; pricing info is incomplete.
How we verified this
Every listed fact about Agent Skils Hub— pricing, features, links, status — is re-read from the vendor's own pages by our verification pipeline and each change is logged with a date; the last pass ran on 30 Oct 2026. Pages mix model-assisted synthesis with human review; see how we verify.
Verification history
We have re-verified Agent Skils Hub 11 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — changed pricing page, summary, pricing tiers, description, who should skip it, features
- — changed pricing page
- — changed who should skip it, description, features, our verdict, summary, pricing tiers, who it suits
- — changed setup time, pricing tiers, use cases, summary, our verdict, features, who it suits, who should skip it, hidden costs, pricing fit, limitations, switching in, switching out, description
- — changed description, our verdict, summary, features, pricing tiers, who it suits, integrations, who should skip it
- — changed pricing page, limitations
Showing the 6 most recent of 11 verification passes.
Free to cite with attribution — this page re-verifies continuously.
How much does Agent Skils Hub cost in 2026?
Agent Skils Hub: Free to use. Pricing last verified 30 Oct 2026.
Plans compared
Every published Agent Skils Hub tier, its price as the vendor publishes it, who it fits, and what it adds over the tier below. The 12-month cost projection follows.
| Plan | Price | Billing | Who it's for |
|---|---|---|---|
| Free | Free | — | Solo developers and small teams who want to vet open-source MCP servers and Claude skills before running them in an agent.Adds: Free entry point: the entire 233,000+ repository catalog, security grades, 0–100 quality scores, Compare, Skill Analyzer and scenario pages, with no paid tier listed. |
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Where the pricing makes sense
The company stage and team size where Agent Skils Hub's pricing actually pencils out — and where peers do it cheaper.
The whole directory is free and MIT-licensed, which puts it below paid alternatives like commercial security scanning vendors or vendor-curated marketplaces. The trade-off isn't subscription cost, it's depth: it's a first-layer scan with no support contract, so teams that need an audited assessment pay for that elsewhere.
Has Agent Skils Hub's pricing changed?
No plan price for Agent Skils Hub has moved since we began logging in Aug 2026; the pricing page is re-read on every verification pass.
Who should pick Agent Skils Hub?
Pick Agent Skils Hub if you are Developers vetting an MCP server or Claude skill before running it in an agent, Security-conscious teams needing a fast risk filter across hundreds of open-source agent tools, Solo builders comparing browser, database or scraping tools for an AI coding stack, Researchers citing agent-skill security and quality data from an open CC-BY-4.0 dataset.
Look elsewhere if you are Teams that need a formal security audit or a vendor-backed guarantee, Buyers who need enterprise contracts, SLAs or a dedicated support contact, Users who want hands-on setup help for individual tools, Anyone expecting every one of the 255,000+ indexed repos to carry a full grade.
Skip it if Skip Agent Skills Hub if you need a signed audit report, vendor-backed SLA or someone to support the tool after you adopt it — this is a rule-based scan over GitHub repos, run by one person as an open-source project.
GeneratedReal-world workflow fit
Concrete scenarios for the personas Agent Skils Hub actually fits — and what changes day-one when you adopt it.
You find a Postgres MCP server on GitHub and want it in your agent today. You look it up on Agent Skills Hub, read the security grade, then check the 0–100 quality score and the README-completeness dimensions before opening the repo.
Outcome: You either rule it out on an UNSAFE grade in under a minute or install it knowing the quality score and exactly which red-flag checks it passed.
Your team has shortlisted a dozen open-source agent tools. You run them through the Skill Analyzer for security grades and platform compatibility, and use Compare to put the finalists side by side.
Outcome: You cut the list to the ones without red flags before anyone spends engineering time on evaluation.
You need browser automation for an agent. You open the browser-use / Playwright MCP scenario page instead of scrolling GitHub search, and get a ranked shortlist scored on quality and community activity.
Outcome: You start evaluating two or three credible options rather than ten, using the quality score to tell finished projects from experiments.
Use cases
- Check the security grade of a Claude skill before you wire it into a production agent.
- Compare MCP servers for browser automation to pick one for your agent pipeline.
- Shortlist quality-scored AI coding assistants for a solo build on a tight timeline.
- Use scenario pages to find database, code review or web scraping tools instead of scrolling GitHub search.
- Track the fastest-growing agent tools through the Monday top-20 star-velocity email.
- Pull the published census dataset to study security and quality trends across the agent-skill ecosystem.
Setup time & first value
How long it actually takes to get something useful out of Agent Skils Hub — broken out by persona, not the marketing-page minute.
No install: open the site and search or browse a category, first useful grade in under two minutes. The Skill Analyzer and Compare view take a few more minutes once you have a shortlist. Reading the methodology page before you rely on an UNSAFE or SAFE verdict is worth ten minutes.
Switching to or from Agent Skils Hub
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From GitHub search: look the repo up on Agent Skills Hub instead, and you get the security grade and quality score layered onto the star data you already had.
- →From a marketplace listing: check the same tool here for red-flag scan results before you install from the marketplace page.
- ↗To a formal security audit: use the scan as triage, then hand the surviving repos to a code-review or audit vendor.
- ↗To GitHub search: go back to star-ranked discovery, accepting that you lose the safety and quality signals.
Integrations
Resources & guides
Tutorials & Learning
YouTube returned 6 videos for “Agent Skils Hub”, and we withheld 6: 6 did not mention Agent Skils Hub. We are showing none, because we could not prove any of them are about Agent Skils Hub.
Official links
What are the best Agent Skils Hub alternatives?
The closest alternatives to Agent Skils Hub we list are Context Mode, Observal and Kasetto. Ranked by direct product-type match, not category overlap.
| Alternative | Pricing | Starts at | Last verified |
|---|---|---|---|
| Context Mode | Freemium | Pricing not published | — |
| Observal | Free | Free | — |
| Kasetto | Free | Free | — |
Alternatives to Agent Skils Hub
View all
Context Mode
Context Mode folds coding-agent tool output before it reaches the model, cutting token spend per request.
Tools that pair well with Agent Skils Hub
Common stack mates teams adopt alongside Agent Skils Hub, with the specific reason each pairing earns its keep.
Context Mode
Context Mode folds coding-agent tool output before it reaches the model, cutting token spend per request.
Observal
Observal is a self-hosted registry and observability layer for AI coding agent skills, MCP servers, and agent traces.
Kasetto
Kasetto syncs skills, commands, MCP servers, and instructions to 23 AI coding agents from one YAML file.
Agent Skils Hub vs its closest rival
Agent Skils Hub's closest head-to-head is with Spider Cloud.
Our Agent Skils Hub vs Spider Cloud verdict written Jul 2, 2026 is more than 90 days old, so it is being re-reviewed against current data and is not shown.
Frequently asked questions
Used Agent Skils Hub? Help shape our editorial sentiment research.