Skip to content

ResearchNVIDIA

NVIDIA SkillSpector: open-source scanner flags vulnerabilities and malicious patterns in AI agent skills

SkillSpector at a glance

NVIDIA's open-source SkillSpector scanner checks AI agent skills for vulnerabilities and malicious patterns before they are installed in tools like Claude Code, Codex and Gemini CLI; research NVIDIA cites found 26.1% of 31,132 analyzed skills contained vulnerabilities and 5.2% showed likely malicious intent.

Ask your AI about this story

Ask what changed in SkillSpector and what to check before using it in your business. Explain how close this is to something I could actually use, and what is still unproven.

Sources checked Primary source (opens in a new tab)

What changed with SkillSpector

NVIDIA's SkillSpector is an open-source security scanner that checks AI agent skills before installation, and the research NVIDIA cites in its documentation shows why: of 31,132 skills analyzed in the study, 26.1% contained at least one vulnerability and 5.2% showed likely malicious intent. The skills run inside coding agents such as Claude Code, Codex and Gemini CLI with implicit trust and minimal vetting.

The study behind the figures, “Agent Skills in the Wild” (Liu and colleagues, January 2026), collected 42,447 skills from two major marketplaces and analyzed 31,132 of them with the SkillScan framework. Data exfiltration (13.3%) and privilege escalation (11.8%) were the most prevalent problems, and skills bundling executable scripts were 2.12 times more likely to contain vulnerabilities than instruction-only skills.

SkillSpector checks a skill against 71 vulnerability patterns across 17 categories, including prompt injection, data exfiltration, privilege escalation, supply-chain risks and MCP tool poisoning. It combines fast static analysis with an optional large language model semantic stage, never executes the skill it scans, and returns a risk score from 0 to 100 with a recommendation; it accepts Git repositories, URLs, zip files, directories and single files.

The scanner is free under the Apache-2.0 license on GitHub, where NVIDIA published the latest release, v2.13.6, on November 16, 2026; SkillSpector also serves as the security tier inside NVIDIA's skill evaluation pipeline, which scans, evaluates and signs skills before publication in the NVIDIA Verified Skills process.

Where to compare SkillSpector

Independent leaderboards and official resources for comparing this model yourself. Rankings change often; each site shows its own current data.

Why SkillSpector matters for businesses

Why it matters: Any business whose developers install third-party agent skills is effectively running unreviewed instructions and code with access to repositories, credentials and internal tools. SkillSpector gives those teams a free pre-install check they can run manually or wire into their pipeline as a gate, so risky skills are flagged before they reach a developer's machine.

Get AI News by email

Verified AI stories, each with its official source, on Monday and Thursday mornings. Add breaking news to hear about the biggest ones as soon as we publish them.

Every story links its official source

Free, in English or Spanish

Unsubscribe with one click, any time

What to send

Sources for this story

  1. Official
  2. Official
  3. Press

Last verified: . IA Compa summarizes public announcements for business readers. We are not affiliated with the companies mentioned. Always confirm details, pricing, and availability with the official source.

  • #nvidia
  • #skillspector
  • #agent-skills
  • #security-scanner
  • #open-source
  • #claude-code
  • #codex
  • #gemini-cli
  • #prompt-injection
  • #supply-chain-security

Want this working in your business?

We turn new AI capabilities into agents, voice receptionists, and automations you can actually use.