Owasp LLM — Agent Skill by Microsoft

From the author

OWASP Top 10 for LLM Applications (2025) knowledge base for identifying, assessing, and remediating large language model security risks.

— the description by Microsoft

AutoGineer summary

Written by AutoGineer from the skill’s own files, not by the author

The owasp-llm skill provides a machine-readable implementation of the OWASP Top 10 for LLM Applications (2025). It serves as a comprehensive reference for developers and security professionals to understand, detect, and mitigate vulnerabilities inherent in large language model systems. Each of the ten identified risks is documented with detailed descriptions, risk assessments, vulnerability checklists, and actionable prevention and remediation strategies.

What it does

Provides a structured index of the 2025 OWASP LLM Top 10 vulnerabilities.
Offers detailed breakdowns for each vulnerability, including risk impact and attack scenarios.
Includes specific vulnerability checklists to help identify system exposure.
Supplies defensive measures and remediation guidance for each security risk.
Categorizes risks into logical groups like Input Integrity, Data Protection, and Access Control.

When to use it

During the design phase of an LLM-integrated application to implement security by design.
When performing a security audit or risk assessment of an existing LLM system.
To train development teams on common LLM-specific security threats.
When developing remediation plans for identified security vulnerabilities in AI workflows.

Advantages

Standardized, industry-recognized framework for LLM security.
Machine-readable format suitable for integration into automated security agents.
Covers both direct and indirect attack vectors, including multimodal risks.
Provides clear, actionable steps for both detection and long-term remediation.

Usage examples

“What are the primary risks associated with Prompt Injection according to the OWASP LLM Top 10?”: The agent retrieves the LLM01:2025 documentation, detailing direct and indirect injection risks, potential consequences like data exfiltration, and a checklist for identifying exposure.
“How can I prevent Excessive Agency in my LLM agent?”: The agent references LLM06:2025, providing guidance on enforcing least privilege, implementing human-in-the-loop requirements for high-risk actions, and defining strict capability boundaries.

Install

Download the skill as a ZIP:
https://mp-staging.autogineer.com/agent-skills/microsoft/owasp-llm/download
Unzip it into the cross-client skills folder of your project:
```
mkdir -p .agents/skills && unzip owasp-llm.zip -d .agents/skills
```

Documentation by Microsoft

OWASP® LLM Top 10 — Skill Entry

This SKILL.md is the entrypoint for the OWASP LLM Top 10 skill.

The skill encodes the OWASP Top 10 for LLM Applications (2025) as structured, machine-readable references that an agent can query to identify, assess, and remediate security risks in large language model systems.

Normative references (LLM Top 10)

Skill layout

SKILL.md — this file (skill entrypoint).
references/ — the LLM Top 10 normative documents.
- 00-vulnerability-index.md — index of all vulnerability identifiers, categories, and cross-references.
- 01 through 10 — one document per vulnerability aligned with OWASP LLM Applications numbering.

Third-Party Attribution

Copyright © OWASP Foundation. OWASP® Top 10 for LLM Applications (2025) content is derived from works by the OWASP Foundation, licensed under CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). Source: https://genai.owasp.org/resource/owasp-top-10-for-llm-applications-2025/ Modifications: Vulnerability descriptions restructured into agent-consumable reference documents with added detection and remediation guidance. OWASP® is a registered trademark of the OWASP Foundation. Use does not imply endorsement.

Files

ATTRIBUTION.mdAdded by AutoGineer973 B
LICENSEAdded by AutoGineer17.9 KiB
SKILL.md2.5 KiB
references/00-vulnerability-index.md3.0 KiB
references/01-prompt-injection.md8.4 KiB
references/02-sensitive-information-disclosure.md6.8 KiB
references/03-supply-chain.md6.9 KiB
references/04-data-and-model-poisoning.md8.0 KiB
references/05-improper-output-handling.md7.6 KiB
references/06-excessive-agency.md6.4 KiB
references/07-system-prompt-leakage.md5.8 KiB
references/08-vector-and-embedding-weaknesses.md6.2 KiB
references/09-misinformation.md6.9 KiB
references/10-unbounded-consumption.md7.4 KiB

ATTRIBUTION.md

Attribution

This folder is an unmodified copy of the Agent Skill "owasp-llm", distributed by AutoGineer (https://autogineer.com/agent-skills).

Source: https://github.com/microsoft/hve-core/tree/f425e943bf228052a23e6bc4810b5faef030314f/.github/skills/security/owasp-llm
Commit: f425e943bf228052a23e6bc4810b5faef030314f
Published by: Microsoft (https://github.com/microsoft)
Publisher: Microsoft (https://github.com/microsoft)
Licence: Creative Commons Attribution-ShareAlike 4.0 International (https://spdx.org/licenses/CC-BY-SA-4.0.html)
Licence evidence: declared in the frontmatter ("license: CC-BY-SA-4.0")
Retrieved: 2026-11-07

The skill's own files are unmodified copies. AutoGineer is not affiliated with or endorsed by the author or the publisher.

Files added by AutoGineer

ATTRIBUTION.md: this file, written by AutoGineer
LICENSE: the SPDX text of CC-BY-SA-4.0 (spdx/license-list-data v3.29.0 (31ba1a50e5397e00a304dbadc76531740e89ee48))

Requirements and compatibility

Written for: GitHub Copilot, Claude Code

Works with: Claude Code, Claude apps, OpenAI Codex, GitHub Copilot, Cursor, Gemini CLI, Kiro, OpenCode, Goose, Amp, Roo Code, JetBrains Junie, Factory Droid, Devin (ex-Windsurf), Mistral Vibe, OpenHands, Letta

“Works with” means the tool reads the open Agent Skills format; a skill written for one tool may use features others lack.

Needs: Claude Code frontmatter

Security

No issues found by our static scan· No issues found

A static scan is not a guarantee. Skills are third-party content: review a skill before you let an agent use it.

Security Audit
CloudflareOfficial
A structured, source-first security audit and vulnerability research framework for codebases and services.
SecurityClaude CodeClaude appsOpenAI Codex+14 more
MITLow riskHas scriptsGitHub stars of the source repository65.3k
Sast Configuration
Seth Hobson· Community
Configure and optimize Static Application Security Testing (SAST) tools like Semgrep, SonarQube, and CodeQL for automated vulnerability detection.
SecurityClaude CodeClaude appsOpenAI Codex+14 more
MITLow riskGitHub stars of the source repository42.3k
Firmware Analyst
Seth Hobson· Community
An expert firmware analyst providing structured guidance for embedded systems security, reverse engineering, and vulnerability research.
SecurityClaude Code
MITLow riskGitHub stars of the source repository42.3k
Policy Enforcer
Seth Hobson· Community
Expert guide for authoring and auditing Cedar authorization policies to secure Claude Code agent tool interactions.
SecurityClaude Code
MITGitHub stars of the source repository42.3k
Backend Development Security Auditor
Seth Hobson· Community
A specialized security auditor for identifying vulnerabilities in backend code and architecture during feature development.
SecurityClaude Code
MITGitHub stars of the source repository42.3k
Memory Forensics
Seth Hobson· Community
Master memory forensics and incident response using the Volatility framework and industry-standard acquisition tools.
SecurityClaude CodeClaude appsOpenAI Codex+14 more
MITLow riskGitHub stars of the source repository42.3k

Download ZIP

The ZIP holds the skill’s files, its licence files and an ATTRIBUTION.md. Download owasp-llm.zip