ai-systems-security

ai-systems-security is a skill for Claude Code from RedHatProductSecurity/prodsec-skills. It costs 41 tokens per session (2,165 once invoked), scanned A, original, Apache-2.0.

A security guide for products that use large language models (LLMs), AI agents, model-serving systems, or MCP tool servers. It focuses on protecting the connections, workflows, tools, and infrastructure around the model.

In plain words
What is it for?
Use it when integrating an LLM API, deploying model-serving infrastructure, building agent workflows, connecting MCP servers, or reviewing an AI system’s security design.
Why use it?
It helps identify security gaps created when AI systems process prompts, call outside services, use tools, or take actions for users.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the prodsec-skills-ge-core plugin — 12 skills shipped together

Good fit Use it when integrating an LLM API, deploying model-serving infrastructure, building agent workflows, connecting MCP servers, or reviewing an AI system’s security design.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/redhatproductsecurity/prodsec-skills/ai-systems-security
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add RedHatProductSecurity/prodsec-skills --skill ai-systems-security
Clone the repo
git clone --depth 1 https://github.com/RedHatProductSecurity/prodsec-skills

Made for: Claude Code.

Or install prodsec-skills-ge-core, the plugin that ships this one along with the rest of its 12 skills.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for ai-systems-security

README.md
[![agentmods](https://agentmods.dev/badge/skills/redhatproductsecurity/prodsec-skills/ai-systems-security/github.svg)](https://agentmods.dev/skills/redhatproductsecurity/prodsec-skills/ai-systems-security)
Your own site
<a href="https://agentmods.dev/skills/redhatproductsecurity/prodsec-skills/ai-systems-security"><img src="https://agentmods.dev/badge/skills/redhatproductsecurity/prodsec-skills/ai-systems-security/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for ai-systems-security

Your own site · 80×15
<a href="https://agentmods.dev/skills/redhatproductsecurity/prodsec-skills/ai-systems-security"><img src="https://agentmods.dev/badge/skills/redhatproductsecurity/prodsec-skills/ai-systems-security.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 41 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,165 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 2 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 7 findings, up to high

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • high Tool Misuse · line 214
    Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).
    Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
  • medium Data Exfiltration · line 123
    Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
    Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
  • medium Data Exfiltration · line 130
    Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
    Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
  • medium Agent Snooping · line 257
    Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.
    Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
  • medium Agent Snooping · line 258
    Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.
    Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
  • medium Agent Snooping · line 259
    Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.
    Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
  • low Privilege Escalation · line 186
    Skill requests more permissions than appear necessary for its stated functionality. Review if elevated access is justified.
    Fix: Request only the minimum permissions required. Document why each permission is needed. Remove broad permissions like '*' or 'all'.
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00041 $0.02165
Opus 5 $0.00020 $0.01082
Sonnet 5 $0.00008 $0.00433
Haiku 4.5 $0.00004 $0.00216

Measured 10d ago against content hash 457245cec112, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

ai-systems-security scanned grade A with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

requests.post(tool_url, headers=headers, json=params)

Runs shell commandslowCapability

Expected in a hook, worth knowing in a rule or an instructions file.

- **Never use `os.system()` or `shell=True`** — always use parameterized APIs
ge-core/ai-systems-security/SKILL.md · 260 lines

How it starts

The opening of the file, as written. The whole thing — 260 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AI Systems Security

Security guidance for teams that serve, integrate, or operate LLMs and AI agents — not for teams training or fine-tuning models. Covers the attack surface introduced when products call LLM APIs, run agent workflows, or expose tool servers.

When to Use

  • Integrating an LLM API (OpenAI, Anthropic, self-hosted) into a product
  • Deploying model-serving infrastructure (OpenShift AI, RHOAI, vLLM)
  • Building or reviewing agent workflows that take actions on behalf of users
  • Connecting MCP servers to agents or LLM-powered features
  • Reviewing AI system architecture for security gaps

When NOT to Use

  • Reviewing AI-generated code for correctness (use ai-code-review)
  • General web application security without AI components (use web-application-security)
  • Container hardening without AI-specific concerns (use container-hardening)

1. Secure LLM Integration

Third-party model providers

  • Obtain information security approval for each provider before integration
  • All connections MUST use TLS — never send prompts or receive responses over unencrypted channels
  • Never hardcode API keys in source code, commit them to version control, or bake them into container images
  • Store credentials in secret management systems (Vault, Kubernetes Secrets) or inject via environment variables at runtime
  • Rotate API credentials on a regular schedule and monitor usage for unexpected patterns
Credential storage:
  FORBIDDEN: Hardcoded in source, committed to git, baked in images
  REQUIRED:  Secret management (Vault, K8s Secrets) or env vars injected at runtime

Self-hosted models

  • Scan models for malicious code before loading — avoid unsafe serialization formats (pickle) that execute code on deserialization; prefer SafeTensors
  • Verify model provenance and integrity (signatures, checksums) before deployment
  • Isolate model-serving processes with dedicated service accounts and restricted network access

2. Prompt Injection Defense

Read the full file on GitHub · 260 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 10d ago First seen · 260 lines · 41 tokens per session scan A 457245cec112

Subscribe to this mod's changes

ai-systems-security is a skill published in the GitHub repository RedHatProductSecurity/prodsec-skills (52 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 41 tokens to every session and 2,165 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 2 findings (makes network calls, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

local-ai-agents

Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…

microsoft/ai-agents-for-beginners · 200 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

insight-error-page

Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…

vercel/next.js · 83 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens

next-partial-prefetching-adoption

Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…

vercel/next.js · 103 tokens