RepoAI
Browse MCPs

Owasp Asi

100TrustedInstructions only

Standardized risk categorization for AI agent vulnerabilities and security findings.

What it does

Adding Owasp Asi to your assistant provides a formal framework for identifying and labeling security risks specific to autonomous agents. It transforms raw incident data or system findings into categorized outputs based on the 2026 industry standards for agentic applications.

By installing this skill, your assistant gains the logic necessary to map technical alerts—such as unauthorized tool usage or memory leaks—to their corresponding ASI identifiers. This allows for consistent reporting and helps prioritize security efforts based on established risk categories.

When it earns its place

Reach for it when…

Apply this skill when you need to audit logs, review security test results, or evaluate potential vulnerabilities within an AI agent ecosystem to ensure they align with industry-standard risk definitions.

Look elsewhere when…

Do not use this for active penetration testing, vulnerability scanning, or automated remediation of security issues, as it only provides classification and mapping rather than active defense mechanisms.

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.

  • classify these security findings according to the OWASP ASI framework
  • map my agent logs to the standard risk categories
  • explain if this vulnerability falls under ASI02 or ASI05
  • what is the severity level of this agent security finding

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.

bash
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/agent-scan/agent_scan/prompt/skills/owasp-asi ~/.claude/skills/owasp-asi

Install 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.

bash
cp -r /tmp/AI-Infra-Guard/agent-scan/agent_scan/prompt/skills/owasp-asi .claude/skills/owasp-asi

If 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

Owasp Asi changes how the work is done. These servers change what Claude can reach while doing it.

How skills and MCP servers differ →