Public story · 2026-03-01 · source-backed

Agent Skills in the Wild: 42,447 Skills Audited (arXiv 2601.10338)

Issue

2026-03-01

Confidence

source-backed

Sources

Redaction

passed

Story

Largest empirical study of MCP/skill ecosystem security. 26.1% of skills contain vulnerabilities spanning 14 patterns. SkillScan detection framework. A "GIF Creator" skill was demonstrated downloading MedusaLocker ransomware.

Anthropic released MCP / Shared entity: MCP / Same source domain / Shared topic / What happened next / Tension
10 actionable skills from today's findings:
Linked by a graph relationship (Anthropic released MCP); both cover MCP; reported by the same outlet (arxiv.org).
Claude Code uses MCP / Shared entity: MCP / Same source domain / Shared topic / What happened next
Every Major MCP Client Is Vulnerable to Prompt Injection via Tool Poisoning. All Seven Tested. All Seven Failed.
Linked by a graph relationship (Claude Code uses MCP); both cover MCP; reported by the same outlet (arxiv.org).
Anthropic released MCP / Shared entities / Shared topic / What happened next
Agent Plugins 1.0.0: the packaging war ended before it started
Linked by a graph relationship (Anthropic released MCP); both cover Agent Skills, MCP; overlapping topics (agent, skill).
Claude Code uses MCP / Shared entity: MCP / Same source domain / Shared topic / Earlier coverage
Skills You Can Learn Today
Linked by a graph relationship (Claude Code uses MCP); both cover MCP; reported by the same outlet (arxiv.org).
Claude Code uses MCP / Shared entity: MCP / Shared topic / What happened next / Tension
The Claude Code Field Guide Hit 53,400 Stars. Here's Why That Matters.
Linked by a graph relationship (Claude Code uses MCP); both cover MCP; overlapping topics (agent, pattern, skill).
Figma supports MCP / Shared entities / Shared topic / Earlier coverage
Anthropic Launches Agent Skills Open Standard — Cross-Vendor Portable Capabilities Are Here
Linked by a graph relationship (Figma supports MCP); both cover Agent Skills, MCP; overlapping topics (agent, skill).
Microsoft supports MCP / Shared entity: MCP / Shared topic / What happened next
MCP Security Debt Hits Critical Mass: 30+ CVEs in 60 Days, PraisonAI CVSS 10, Azure AI Foundry CVSS 10
Linked by a graph relationship (Microsoft supports MCP); both cover MCP; overlapping topics (agent, ecosystem, framework, pattern, security).
Claude Code uses MCP / Shared entity: MCP / Same source domain / Shared topic / What happened next
Static analysis catches malicious agent skills at 0.93 AUC and misses shell-command host destruction 100% of the time.
Linked by a graph relationship (Claude Code uses MCP); both cover MCP; reported by the same outlet (arxiv.org).

Source trail

26.1% of skills contain vulnerabilities

Entities

Agent Skills GIF Creator Largest MCP MedusaLocker Skills Audited SkillScan Wild

Provenance

Canonical issue: Ramsay Research Agent — 2026-03-01
AI generated: no
Story unit: 2026-03-01-agent-skills-in-the-wild-42-447-skills-audited-arxiv-2601-10338
Labels: source-backed, canonical briefing excerpt

Fetching from the wire…

Agent Skills in the Wild: 42,447 Skills Audited (arXiv 2601.10338)

Story

Related stories

10 actionable skills from today's findings:

Every Major MCP Client Is Vulnerable to Prompt Injection via Tool Poisoning. All Seven Tested. All Seven Failed.

Agent Plugins 1.0.0: the packaging war ended before it started

Skills You Can Learn Today

The Claude Code Field Guide Hit 53,400 Stars. Here's Why That Matters.

Anthropic Launches Agent Skills Open Standard — Cross-Vendor Portable Capabilities Are Here

MCP Security Debt Hits Critical Mass: 30+ CVEs in 60 Days, PraisonAI CVSS 10, Azure AI Foundry CVSS 10

Static analysis catches malicious agent skills at 0.93 AUC and misses shell-command host destruction 100% of the time.

Source trail

Entities

Provenance

Agent Skills in the Wild: 42,447 Skills Audited (arXiv 2601.10338)

Story

Related stories

10 actionable skills from today's findings:

Every Major MCP Client Is Vulnerable to Prompt Injection via Tool Poisoning. All Seven Tested. All Seven Failed.

Agent Plugins 1.0.0: the packaging war ended before it started

Skills You Can Learn Today

The Claude Code Field Guide Hit 53,400 Stars. Here's Why That Matters.

Anthropic Launches Agent Skills Open Standard — Cross-Vendor Portable Capabilities Are Here

MCP Security Debt Hits Critical Mass: 30+ CVEs in 60 Days, PraisonAI CVSS 10, Azure AI Foundry CVSS 10

Static analysis catches malicious agent skills at 0.93 AUC and misses shell-command host destruction 100% of the time.

Source trail

Entities

Provenance