Every premium and members-only resource for $9.99/month, as a no-commitment subscription. See the plans

PalateKit

Sign in
Menu

Request a resource

Describe the resource you are missing in one sentence. We will email you when it is published.

0 / 280

A page, a video or an image online.

Unlock PalateKit

Unlock everything

Plan

Paying means accepting our terms of sale.

Secure payment by Stripe. Receipts and invoices sent by email.

Welcome back

Sign in to install, download or save this resource. A free account is enough for free resources.

Harden an AI deployment: injection, leaks, model theft v1.0.0

Use this resource to harden an LLM deployment against injection, data leaks and model theft

New

  • 326 downloads
  • 192 likes
  • 136 saves

Free

Install with the CLI

In a terminal, from the folder you want:

npx --yes palatekit@latest install ai-llm-security-controls

With no option, the skill goes to ~/.agents/skills/; add --agent claude-code (or codex, cursor…) to target an agent.

First time: connect the CLI with

npx --yes palatekit@latest login

The agent builds an AI-specific threat model, then adds NeMo guardrails, output filtering with personal data scrubbing, API limits and model weight verification. You get controls to deploy and compliance checkpoints for SOC 2 or ISO 27001.

Mostly picked by

Updated November 2, 2026

Card for the “Harden an AI deployment: injection, leaks, model theft” skill

Details

Works in
Install guides
Generates with
Your agent's model
Licence
MIT

Similar resources

View all

What you get

  • SKILL.md : AI threat model and prompt injection defense
  • SKILL.md : NeMo guardrails, output filtering and LLM API security

Good fit for

  • Your AI handles sensitive user data
  • You must bring your AI in line with SOC 2 or ISO 27001

Not for

    Tags

    • ai
    • security
    • guardrails
    • compliance

    About

    You say

    “Harden our LLM API: per-key limits, personal data filtering and guardrails.”

    Card for the “Harden an AI deployment: injection, leaks, model theft” skill

    Skill published by bagelhole under the MIT license, delivered in its original version.

    Show more

    How to use it

    1. Install the skill in your agent's skills folder
    2. Ask your agent to harden an LLM deployment against injection, data leaks and model theft
    3. Review and adjust the result

    Content preview

    Files

    • LICENSE1.1 KB
    • NOTICE-PALATEKIT.md0.5 KB
    • SKILL.md10.3 KB

    Origin and licence

    Listed unmodified

    Author
    bagelhole
    Source
    github.com/bagelhole/devops-security-agent-skills/tree/0365f57a079b1332f95cf26e31dd2d5332a8399f/security/ai/ai-security-hardening
    Licence
    MIT

    Notice

    « ai-security-hardening » (bagelhole, MIT) livré dans sa version d'origine : fiche en français et en anglais par PalateKit.

    User reviews

    –

    No ratings yet

    Loading reviews…

    Delete your review?

    Your rating and comment will be removed from the page.