Edgeone Skill Scanner
88TrustedInstructions onlyStatic analysis for agent instructions to identify potential security vulnerabilities.
What it does
Edgeone Skill Scanner adds a vetting layer to your agent setup by reviewing the content of instruction files. It performs local, static analysis to detect suspicious patterns or risks, ensuring you understand the implications of the logic contained within any given skill file before you rely on it.
This utility operates entirely on your local machine. It does not transmit file contents or personal data to any external server, maintaining strict data privacy while it processes the syntax and instructions of your installed agent capabilities.
When it earns its place
Reach for it when…
Apply this scanner whenever you import a new skill from an external repository or community source. It is designed for individual users who want to perform due diligence on instruction sets that dictate how their AI agent manages files, interacts with system processes, or executes automated tasks.
Look elsewhere when…
Do not rely on this for comprehensive threat detection or active defense against malicious binaries, as it is limited to static analysis of instruction text. It cannot verify the runtime behavior of external applications, plugins, or compiled software, and it is not a substitute for standard system-level antivirus software or network security monitoring.
What to say to use it
You never invoke a skill directly — the assistant picks it up when your request matches what it covers. These are the kinds of things that do that.
- run a safety check on this skill
- is this agent skill safe to use
- scan all my installed skills for security
- analyze the security risks in this skill directory
Install it
Install for all your projects
Copy the skill directory into your personal skills folder. Claude picks it up on its own the next time your request matches what it does — there is nothing to enable.
git clone --depth 1 --branch main https://github.com/Tencent/AI-Infra-Guard.git /tmp/AI-Infra-Guard
mkdir -p ~/.claude/skills
cp -r /tmp/AI-Infra-Guard/skills/edgeone-skill-scanner ~/.claude/skills/edgeone skill scannerInstall for one project only
Put it in the project's own skills folder instead, so it can be committed and shared with everyone working on that repository.
cp -r /tmp/AI-Infra-Guard/skills/edgeone-skill-scanner .claude/skills/edgeone skill scannerIf a skill of the same name exists in both places, the personal one wins.
Installing on Claude.ai or through the API instead? The full guide covers every surface.
Pair it with an MCP server
Edgeone Skill Scanner changes how the work is done. These servers change what Claude can reach while doing it.
- Playwright MCP Server46The skill decides what is worth testing; Playwright MCP Server is what drives the browser to test it.
- GitHub MCP Server53With GitHub MCP Server, review guidance can be applied to the actual diff rather than to code pasted into the conversation.
- Circleci MCP Server58Circleci MCP Server surfaces which builds failed, giving the skill's review something concrete to work from.