Skip to main content
skill CX Codex

llm-input-output-guardrail-middleware-development — Claude Code skill

llm-input-output-guardrail-middleware-development is a Claude Code skill for designing protective middleware around LLM agents. It guides work on untrusted inputs, tool permissions, and outbound data checks.

no license ★ 73 updated 2026-11-15

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add W4L33Dx/Cyber-Monster --skill llm-input-output-guardrail-middleware-development -a claude-code
Open on GitHub

What llm-input-output-guardrail-middleware-development does

The skill lays out a middleware design for Python and TypeScript applications using LLMs, MCP, or agents. It keeps system instructions distinct from untrusted content, checks tool calls against JSON Schema, allowlists, and caller permissions, and adds source tags and canary tokens to retrieved RAG fragments. Outbound text and tool arguments are checked for personal data, secrets, and URLs.

The guide includes a validation sequence, test-case examples, and a self-check command: python -m py_compile guardrail.py. Its specified stack uses Python 3.11 and jsonschema for JSON Schema 2020-12 validation. Unknown tools or arguments are rejected by default.

What it helps with

  • separate system instructions from retrieved RAG content
  • validate tool-call arguments against JSON Schema
  • restrict tool access by allowlist and caller permissions
  • filter secrets and personal data from outbound content

Example prompts for the agent

  • Design middleware that separates system instructions from untrusted RAG fragments.
  • Add JSON Schema checks and argument allowlists for MCP tool calls.
  • Draft outbound filters for secrets, personal data, and disallowed URLs in agent responses.

How to install llm-input-output-guardrail-middleware-development

1Skills CLI (Vercel)

npx skills add W4L33Dx/Cyber-Monster --skill llm-input-output-guardrail-middleware-development -a claude-code

2Manually: put the files in place

~/.claude/skills/llm-input-output-guardrail-middleware-development/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/llm-input-output-guardrail-middleware-development/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about llm-input-output-guardrail-middleware-development

How do I install llm-input-output-guardrail-middleware-development in Claude Code?

Run npx skills add W4L33Dx/Cyber-Monster --skill llm-input-output-guardrail-middleware-development -a claude-code. The skill file is at ~/.claude/skills/llm-input-output-guardrail-middleware-development/SKILL.md.

What does llm-input-output-guardrail-middleware-development do?

It guides the design of middleware that separates instructions from untrusted data, validates tool calls, and filters outbound content.

What does llm-input-output-guardrail-middleware-development require?

The described stack uses Python 3.11 and jsonschema to validate against JSON Schema 2020-12.

Similar skills

All skills for Security →

From the same set61

The whole set →
memory-forensics-network-and-credential-artifacts memory-forensics-network-and-credential-artifacts is a Claude Code skill for examining network data and credentials in Windows memory images. It activates for memory-artifact analysis and credential recovery. CX Codex +1 ★ 73repo stars legal-and-regulatory-notification legal-and-regulatory-notification is a Claude Code skill that maps incident disclosure obligations, statutory clocks, and counsel evidence. Activates when breaches trigger reporting duties. CX Codex +1 ★ 73repo stars edr-api-triage-and-containment-scripting edr-api-triage-and-containment-scripting is a Claude Code skill for building CLI tools around EDR investigations. It activates for scripting host triage and controlled isolation. CX Codex +1 ★ 73repo stars sql-injection sql-injection is a Claude Code skill that evaluates untrusted inputs reaching database statements. Runs during audits of login parameters, report filters, and sorting clauses. CX Codex +1 ★ 73repo stars ics-and-ot-passive-network-monitoring ics-and-ot-passive-network-monitoring is a Claude Code skill for passive traffic inspection in OT and SCADA setups. It inspects Modbus, S7comm, and DNP3 exchanges without active device polling. CX Codex +1 ★ 73repo stars syslog-and-cloudtrail-stream-parsers syslog-and-cloudtrail-stream-parsers is a Claude Code skill for building streaming parsers for cloud audit logs and syslog. It applies to CloudTrail, Azure Activity/SignIn, GCP Audit Logs, RFC 5424, and CEF feeds. CX Codex +1 ★ 73repo stars email-security-monitoring email-security-monitoring is a Claude Code skill that inspects email headers to uncover spoofing, phishing, and BEC threats. Triage workflows trigger it on suspicious inbox messages. CX Codex +1 ★ 73repo stars container-and-kubernetes-forensics container-and-kubernetes-forensics is a Claude Code skill for investigating container and Kubernetes incidents. It examines container filesystems, image layers, audit logs, and node evidence. CX Codex +1 ★ 73repo stars

Project sponsors

Companies that keep this analytics open