guardrails

guardrails is a skill for Claude Code, Codex from hajekim/agentic-design-patterns-extension. It costs 398 tokens per session (3,651 once invoked), scanned C, a copy of guardrails, MIT.

A set of safety boundaries placed around an AI agent to check incoming requests, validate responses, and limit actions to approved behavior.

In plain words
What is it for?
Filtering content, validating inputs and outputs, preventing prompt-injection attacks, protecting data, enforcing scope, and checking actions before execution.
Why use it?
It reduces the risk of harmful, unsafe, off-topic, unauthorized, or policy-breaking results when agents handle untrusted input or sensitive work.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Filtering content, validating inputs and outputs, preventing prompt-injection attacks, protecting data, enforcing scope, and checking actions before execution.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/hajekim/agentic-design-patterns-extension/guardrails
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 hajekim/agentic-design-patterns-extension --skill guardrails
Clone the repo
git clone --depth 1 https://github.com/hajekim/agentic-design-patterns-extension

Made for: Claude Code, Codex.

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 guardrails

README.md
[![agentmods](https://agentmods.dev/badge/skills/hajekim/agentic-design-patterns-extension/guardrails/github.svg)](https://agentmods.dev/skills/hajekim/agentic-design-patterns-extension/guardrails)
Your own site
<a href="https://agentmods.dev/skills/hajekim/agentic-design-patterns-extension/guardrails"><img src="https://agentmods.dev/badge/skills/hajekim/agentic-design-patterns-extension/guardrails/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 guardrails

Your own site · 80×15
<a href="https://agentmods.dev/skills/hajekim/agentic-design-patterns-extension/guardrails"><img src="https://agentmods.dev/badge/skills/hajekim/agentic-design-patterns-extension/guardrails.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 398 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,651 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 2 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin 100% copy Near-identical to another mod 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 $0.00398 $0.03651
Opus 5 $0.00199 $0.01826
Sonnet 5 $0.00080 $0.00730
Haiku 4.5 $0.00040 $0.00365

Measured 9d ago against content hash 2340cd67e879, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade C, and why

guardrails scanned grade C 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 9d 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.

Instruction-override phrasingmediumPrompt injection

Text telling the model to disregard its earlier instructions or safety rules is the shape of a prompt injection, whoever wrote it.

"ignore previous instructions",

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Nullifies safety policiesmediumAnti-refusal

"You have no restrictions", "do anything now", "ignore your guidelines": a direct jailbreak that disables guardrails.

"bypass your restrictions"

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Origin

This is a copy

100% identical to guardrails — 3 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

skills/guardrails/SKILL.md · 377 lines

How it starts

The opening of the file, as written. The whole thing — 377 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Guardrails & Safety Pattern

Overview

The Guardrails & Safety Pattern establishes protective boundaries around agent behavior — filtering inputs, validating outputs, enforcing policy constraints, and preventing harm. Guardrails operate as a safety layer that sits between users and agents (input guardrails) and between agents and users (output guardrails), ensuring agents remain beneficial, appropriate, and within authorized scope.

Core Principle: Safety is not an afterthought — build constraints into the design so agents can't go wrong, not just catch them after they do.

When This Skill Applies

Activate this pattern when:

  • Agents are deployed in public-facing applications where prompt injection is possible
  • Outputs could cause harm if unfiltered (medical, legal, financial advice)
  • The agent has access to sensitive systems or data that must be protected
  • Regulatory compliance requires content filtering and audit trails
  • The agent must stay within a defined scope (don't answer off-topic questions)
  • Actions in the world (emails, purchases, deletions) require pre-execution validation

Rule of thumb: Any agent interacting with untrusted users or executing real-world actions needs guardrails — the question is which ones, not whether.

Guardrail Types

Input Guardrails

  • Topic filtering: Block requests outside the agent's scope
  • Injection detection: Detect prompt injection attacks
  • PII detection: Identify and handle personal information
  • Content moderation: Filter harmful, offensive, or inappropriate requests
  • Rate limiting: Prevent abuse through request throttling

Output Guardrails

  • Fact checking: Verify claims against authoritative sources
  • Toxicity filtering: Remove harmful language from responses
  • Hallucination detection: Flag responses that may not be grounded
  • PII redaction: Remove sensitive data from outputs
  • Scope enforcement: Ensure responses stay within authorized domain

Read the full file on GitHub · 377 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. 9d ago First seen · 377 lines · 398 tokens per session scan C 2340cd67e879

Subscribe to this mod's changes

guardrails is a skill published in the GitHub repository hajekim/agentic-design-patterns-extension (1 stars, last pushed 5mo ago), licensed MIT. It adds 398 tokens to every session and 3,651 once invoked, about $0.0020 per session on Opus 5. A static security scan graded it C with 2 findings (instruction-override phrasing, nullifies safety policies). It is 100% identical to guardrails, differing in 3 lines, and is treated as a copy.