✍️Writing & Content35🎨Image Generation43🎬Video & Animation76🎵Audio & Music67💬Chatbots & Assistants58💻Coding & Development284📈Marketing & SEO87Productivity218🎯Design & UI/UX71📊Data & Analytics69📚Education & Research30💼Business & Finance75🏥Healthcare & Wellness18🔍Search & Knowledge17🤖AI Agent Infrastructure121🛡️AI Security & Testing17🧊3D & Spatial21🔎SEO Tools34🏡Real Estate4🗃️Data Extraction32🧠ADHD & Focus Tools9
Listed in AI Security & Testing with 17 other toolsPart of 1554+ curated AI tools on AISO
MCP Skills logo

MCP Skills

Trust scores and monitoring for MCP servers and agent skills, checked before you install

0
freemiumFree tier: 10 scans/day with compact responses, no API key needed. Single full report $2. Developer Pro $19/mo or $149/yr adds monitoring and an API key. Higher published tiers at $29/mo and Team $99/mo ($249/yr option listed).View full pricing →

Visit MCP Skills

https://mcpskills.io

About MCP Skills

MCP Skills is a pre-install trust layer for MCP servers and agent skills. It turns public source, package, vulnerability, and supply-chain data into scored pages, trust badges, monitoring, and API workflows you can consult before an unknown tool reaches an agent. The reason it exists is documented on the site: OX Security submitted a benign proof-of-concept malicious MCP server to eleven public registries in April 2026, and nine of them — including LobeHub and Cursor Directory — published it with no security review. Only GitHub rejected it. MCP Skills scores each repository across fifteen trust signals grouped into four dimensions and assigns a tier: Verified means a composite of 7.0 or higher with dimension floors and no disqualifiers, Established means solid with caveats worth checking, and New means promising but unproven. It had scored 2,631 repositories across five registries as of late July 2026, and its own research found 1,184 malicious skills in the ClawHavoc incident and that 36.82% of skills carried at least one security flaw. Maintainers can claim a gold Verified badge on their score page or embed a live trust badge. Access is layered: a free tier allows ten scans per day with compact responses, a single full report costs $2, and Developer Pro adds monitoring and API keys.

Key Features

15 trust signals across 4 dimensions scored per repository
Verified / Established / New tiering with explicit thresholds and disqualifiers
2,631 repositories scored across 5 registries
Safety Lab for running deeper checks on a server
Embeddable live trust badges and claimable Verified badges for maintainers
Monitoring so a repo's score change alerts you after install
API access for wiring trust checks into workflows
Published original research on malicious skills and ecosystem flaw rates

MCP Skills Pros & Cons

Pros

  • +Addresses a demonstrated gap — nine of eleven registries published a malicious PoC server unreviewed
  • +$2 for a single full report is a low bar to check one questionable server
  • +Free tier at 10 scans/day is enough for occasional personal use
  • +Scoring methodology and thresholds are published rather than opaque

⚠️ Cons

  • Scores are heuristic signals, not a security audit, and can be gamed over time
  • Coverage of 2,631 repos leaves plenty of long-tail servers unscored
  • Several overlapping paid tiers make the pricing harder to reason about than it should be
  • A trust badge creates a false-confidence risk if users stop reading the details

Who Is MCP Skills Best For?

👤Developers installing MCP servers from registries with no security review
👤Teams that need a documented check before adding agent tooling
👤MCP maintainers who want a verifiable trust badge for their server

Tags

mcpsupply-chainsecuritytrust-scoresmonitoringapi
🏷️

Is this your tool?

Claim your listing to get a Featured badge, edit your description, and stand out from competitors. All plans include a permanent dofollow backlink to your site.

Claim Now →

ChatGPT already recommends MCP Skills. Does it recommend yours?

If you're building in AI Security & Testing, run a free AI-visibility scan on your own product — we ask ChatGPT across 5 prompt angles and score how often you get named. ~30 seconds, no signup, no card.

Stay updated on AI Security & Testing tools — join our weekly newsletter

One concise email with fresh launches, trending picks, and featured standouts.

Alternatives to MCP Skills

View all MCP Skills alternatives →

Agent connectivity: not yet verified