skills.rest
Tencent
TencentOFFICIAL
Explore all skills by this author →
WebsiteGitHub

owasp-asi

Classify AI agent security findings using the OWASP ASI 2026 framework.

4.4k|438|Updated Dec 25, 2024
Search Tags:#owasp#vulnerability management#ai security#risk classification#agentic applications#asi
One-click install
npx skills add https://github.com/Tencent/AI-Infra-Guard --skill owasp-asi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: owasp-asi
Source: https://github.com/Tencent/AI-Infra-Guard/tree/main/agent-scan/prompt/skills/owasp-asi
Command: npx skills add https://github.com/Tencent/AI-Infra-Guard --skill owasp-asi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a standardized framework for classifying security findings related to AI agent applications, aligning them with the OWASP Top 10 for Agentic Applications (ASI) 2026.

Core Features & Use Cases

  • Risk Categorization: Maps security findings to specific ASI risk types (e.g., Agent Goal Hijack, Tool Misuse).
  • Detection Source Mapping: Correlates various detection methods (e.g., prompt injection tests, code audit) to relevant ASI categories.
  • Severity Classification: Assigns severity levels (Critical, High, Medium, Low) to findings with clear remediation timelines.
  • Use Case: After running an AI security scan, use this skill to categorize the identified vulnerabilities, such as "API keys exposed" being mapped to ASI06 (Memory & Context Poisoning) and classified as High severity.

Quick Start

Load the owasp-asi skill to classify security findings according to the OWASP ASI framework.

Frequently Asked Questions about owasp-asi

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I classify AI agent security findings using the OWASP ASI framework?▼

AI agent security findings are classified by mapping vulnerabilities to specific OWASP ASI risk types, such as Agent Goal Hijack, and assigning severity levels from Critical to Low for clear remediation timelines.

What is the OWASP Top 10 for Agentic Applications used for in vulnerability management?▼

The OWASP Top 10 for Agentic Applications provides a standardized framework for categorizing security risks specific to AI agents, ensuring consistent risk assessment and facilitating targeted remediation during security audits.

How do I map prompt injection tests to ASI risk categories?▼

Prompt injection tests map to relevant ASI categories by correlating the detection source with standardized risk types like Memory and Context Poisoning, then assigning an appropriate severity level for remediation.

Can I use this OWASP ASI classification for vulnerabilities found during code audits?▼

Yes, detection sources from code audits are mapped to relevant ASI categories like Tool Misuse. Findings are then assigned severity levels to establish clear remediation timelines for AI security audits.

What is the best way to standardize severity levels for AI security vulnerabilities?▼

Standardizing severity levels for AI security vulnerabilities involves assigning findings to Critical, High, Medium, or Low categories based on the OWASP ASI framework, which directly dictates remediation timelines.

Related Skills

Explore More Software Engineering Skills→
tbartel74tbartel74
14

security-audit-scanner

Automate OWASP Top 10 checks, secret detection, and dependency audits for Vigil Guard services.

Community
Advanced
gtpgg1013gtpgg1013

code-review

Review git diffs for security, quality, and maintainability issues.

Community
Advanced
uphiagouphiago
1.1k

hunt-llm-ai

Identify exploitable LLM/AI vulnerabilities crossing trust boundaries with validation gates.

Community
Advanced
RepairYourTechRepairYourTech
6

security-scanning-security-hardening

Orchestrate multi-agent security hardening across application, infrastructure, and compliance domains.

Community
Advanced