Lijithvmv

Guard Layer — AI skill for Claude Code

AI community

Runtime guardrails between an AI agent and its tools: scan what it reads, gate what it does, and track the session so a secret read earlier can't be exfiltrated later.

How to install Guard Layer

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Lijithvmv/Guard-Layer and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Guard Layer does

Runtime guardrails between an AI agent and its tools: scan what it reads, gate what it does, and track the session so a secret read earlier can't be exfiltrated later. Containment over detection — pure Python, zero dependencies, ~0.2 ms per tool call.

Alternatives in AI

  • Hol Guard — Open-source antivirus for AI agents: block risky tools, secret access, prompt injection, malicious packages, M 485 ★
  • Prismor — Self-hosted runtime control plane for AI agents 334 ★
  • Hazmat — Containment for AI agents - user isolation, sandboxed execution, network controls, backup/rollback 167 ★

README

GuardLayer — runtime guardrails between an agent and its tools

GuardLayer

Stop an AI agent from doing harm after it reads something an attacker wrote.

GuardLayer sits between an agent and its tools. It checks what the agent **reads**, decides whether what it is about to **do** may run, needs a human, or is refused, and remembers what the session has already seen, so an action is judged in context. Pure Python, zero dependencies, about 0.2 ms per tool call in-process.

![Python](https://img.shields.io/badge/python-3.10–3.13-blue) ![License](https://img.shields.io/badge/license-MIT-green) ![Status](https://img.shields.io/badge/status-beta-yellow) ![Dependencies](https://img.shields.io/badge/core%20dependencies-0-brightgreen)

**Documentation: https://lijithvmv.github.io/Guard-Layer/**

Why this approach

Any text an agent reads (a web page, an email, an issue, a tool result) can carry instructions, and the model can't reliably tell them apart from yours. Detectors help, but on attacks they have never seen they catch a minority (our own held-out numbers are below). So GuardLayer does not bet on detection. It stops the **harm**:

  1. What the agent reads is scanned for injections, secrets and personal data, and the session remembers it: untrusted content, an injection, sensitive data.
  2. What the agent is about to do goes through a tool policy: destructive commands, credential files, exfiltration endpoints, and your own argument rules.
  3. The two meet in the session: a secret read earlier and now being sent out is blocked; any side effect after the agent read an injection needs a human; untrusted content plus sensitive data plus a network call needs a human. This works even when the injection itself was never detected.
  4. Every decision goes to a tamper-evident audit log.

Quickstart: Claude Code

# pilot.toml: record what would happen, enforce nothing