SkillSpector

What is SkillSpector?

SkillSpector is a security scanner for AI agent skills, published by NVIDIA, that helps you decide whether a skill should be installed. Skills can look harmless while hiding risky instructions, overbroad permissions or code that does more than its description says. SkillSpector checks 68 vulnerability patterns across 17 categories, among them prompt injection, data exfiltration, privilege escalation, supply-chain issues, tool misuse and MCP tool poisoning. Targets can be Git repositories, URLs, zip archives, folders or single files, and runs fast static checks by default. An optional LLM pass checks whether a skill's code really matches the behavior it advertises. Known vulnerable dependencies are looked up through OSV.dev, with an offline fallback. Output can be terminal, JSON, Markdown or SARIF for CI. You clone it from GitHub and run it yourself, and no pricing is listed. It also forms Tier 1 of NVIDIA's SkillEvaluator.

Use cases

• Scan a skill before installing it
• Detect prompt injection in skills
• Find hidden instructions and tool poisoning
• Flag vulnerable skill dependencies
• Add skill scanning to CI
• Review skills before publishing

Who it's for

• Skill authors
• Security teams
• Agent developers

Why use SkillSpector

• Scanning before install catches risky instructions and code that a skill's description does not mention.
• Static mode is fast and deterministic, so it suits local iteration and automated checks.
• SARIF and JSON output feed CI pipelines and code scanning dashboards.
• The triage guidance tells you which findings to block, fix or document.
• Semantic mode flags skills whose stated purpose disagrees with what the code does.

SkillSpector features

• 68 vulnerability patterns in 17 categories
• Scans Git repos, URLs, zip files, directories and files
• Static checks that run without an LLM
• Optional LLM semantic analysis of description versus behavior
• Live dependency lookups through OSV.dev with offline fallback
• Output as terminal, JSON, Markdown or SARIF
• MCP least-privilege and tool-poisoning checks
• YARA signature and taint-tracking checks
• Runs as the security tier of SkillEvaluator

SkillSpector integrations

• OSV.dev
• OpenAI
• SkillEvaluator
• GitHub

How to get started with SkillSpector

1. Clone the SkillSpector repository and run make install inside a Python virtual environment.
2. Run skillspector scan against a directory, SKILL.md, Git URL or zip file.
3. Pass the no-llm flag for static-only review, or set a provider and API key for semantic analysis.
4. Choose terminal, JSON, Markdown or SARIF output for your workflow.
5. Triage the findings and block, fix or document each issue before release.

SkillSpector limitations

• Semantic analysis needs an LLM provider and API key
• Scan results are one release gate and do not replace evaluation
• Installed from source with Python, not as a hosted service

SkillSpector pricing

SkillSpector is open source, so you can use and self-host it for free.

Prices can change and may vary by country or exclude tax. Check the official pricing page.

SkillSpector alternatives

The main SkillSpector alternatives listed on Debuly are Semgrep, Promptfoo, Repo Forensics, ClawHub, Agent Governance Toolkit and Agent Skills. The table compares pricing and platforms; order is A–Z.

1. 

   Agent Governance Toolkit

   Microsoft's open-source toolkit for enforcing policy, identity, sandboxing and audit records on autonomous AI agents before they reach production.

   Developer platformsAutomation and agents
   Open sourceVisit Agent Governance Toolkit (official site)
2. 

   Agent Skills

   Open standard for packaging instructions, scripts and references in a SKILL.md folder that AI agents load on demand.

   Developer platformsAutomation and agentsCoding
   Open sourceVisit Agent Skills (official site)
3. 

   ClawHub

   Public registry where you find, install and publish skills and plugins for the OpenClaw AI assistant, with a CLI and API.

   Developer platformsAutomation and agentsCoding
   Open sourceVisit ClawHub (official site)
4. 

   Promptfoo

   Open-source LLM evaluation and red-teaming tool that finds prompt injections, jailbreaks and data leaks in AI apps and agents.

   Developer platformsCodingAutomation and agents
   Open sourceVisit Promptfoo (official site)
5. RF

   Repo Forensics

   Offline security scanner that audits AI-agent repos, skills, plugins and MCP servers for malicious code before you install them.

   Developer platformsCoding
   Open sourceVisit Repo Forensics (official site)
6. 

   Semgrep

   Application security platform for static analysis, supply chain and secrets scanning, with AI-assisted triage and fixes for developers and coding agents.

   Developer platformsCoding
   FreemiumVisit Semgrep (official site)

SkillSpector FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: from docs.nvidia.com · fetched 24 Oct 2026 · Vendor trademark, used only to identify the product; removed on request
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.