Works with the AI Agents you already use
github policy guard
Enforce GitHub Acceptable Use Policies and prevent malware/PUA patterns in AI-generated system scripts and utilities.
$5
github policy guard
Example session with this skill installed
Write a PowerShell script to optimize network latency by changing the TCP Global Parameters in the registry. Include a way to undo it.
- Read your context and instructions
- Compiled the github policy guard
I have generated a policy-compliant PowerShell script for network optimization. Following the security guardrails, I have avoided Invoke-Expression, implemented a dry-run mode, and included a deterministic restore function.
Network Optimization Script (Compliant)
<#
.SYNOPSIS
Configures QoS NonBestEffortLimit via documented registry keys.
Complies with GitHub Policy Guard requirements for idempotence and safety.
#>
param(
[switch]$DryRun,
[switch]$Restore
)
$RegPath = "HKLM:\SOFTWARE\Policies\Microsoft\Windows\Psched"
$Name = "NonBestEffortLimit"
$BackupFile = "tcp_backup_$(Get-Date -Format 'yyyyMMdd_HHmmss').reg"
function Restore-Defaults {
Write-Host "Restoring standard Windows defaults..."
if (Test-Path $RegPath) {
Remove-ItemProperty -Path $RegPath -Name $Name -ErrorAction SilentlyContinue
}
Write-Host "Restore complete."
}
if ($Restore) {
Restore-Defaults
exit
}
if ($DryRun) {
Write-Host "[DRY RUN] Would set $RegPath\$Name to 0"
exit
}
# Create Backup
reg.exe export $RegPath $BackupFile /y
# Idempotent Mutation
if (!(Test-Path $RegPath)) {
New-Item -Path $RegPath -Force
}
$args = @("add", $RegPath, "/v", $Name, "/t", "REG_DWORD", "/d", "0", "/f")
& reg.exe @args
Write-Host "Optimization applied. Backup saved to $BackupFile"
Next steps
- Run with
-DryRunto verify the registry path on your specific Windows build. - Ensure you run the terminal as Administrator, as
HKLMmodifications require elevated privileges. - Test the
-Restoreflag immediately after application to verify the rollback mechanism.
Connects securely to your tools. The creator never sees your data.
About this skill
Your AI agent might be one bad script away from getting your GitHub account suspended. Modern coding agents often default to "stealth" execution patterns or hardcoded secrets that trigger automated malware and PUA (Potentially Unwanted Application) detection systems.
What it does
- Enforces policy compliance by blocking code patterns that violate GitHub's Acceptable Use Policies regarding malware and exploits.
- Prevents stealth execution by banning hidden window flags and concealed command prompts in VBScript, PowerShell, and Bash.
- Hardens secret management by automatically adding cryptographic keys and password files to
.gitignoreand enforcing environment variable usage. - Validates system mutations by requiring idempotence, dry-run flags, and rollback mechanisms for any script modifying the registry or network.
- Secures release workflows by ensuring binaries are built via GitHub Actions with SLSA provenance rather than committed directly to the repo.
How it works
- The agent monitors any turn involving Windows utilities, system scripts (Rust, Go, PowerShell, Bash), or repository packaging.
- It scans generated code for prohibited patterns like
Invoke-Expression, hardcoded user paths, orWScript.Shellhidden flags. - It applies mandatory guardrails such as automatically configuring
.gitignorefor certs and requiring parameterized execution for system commands. - It runs a final verification checklist to ensure no secrets are staged and all architectural safety requirements are met before the task is marked complete.
Frameworks & tools
The skill supports Windows systems programming, PowerShell, Bash, Rust, Go, Python, and C++. It specifically handles compliance for GitHub Actions, Git repositories, and installer frameworks like NSIS and InnoSetup.
Why this beats prompting it yourself
Prompting "be secure" is too vague to stop an agent from using a convenient but dangerous eval() or Invoke-Expression. This skill provides a specific technical taxonomy of MITRE ATT&CK patterns and GitHub-specific prohibitions that most LLMs will overlook in favor of functional code.
Use cases
- Developing Windows system utilities without triggering Defender or SmartScreen malware alerts.
- Automating repository setup with pre-configured
.gitignoreand security policy documentation. - Writing safe installers that include guarded uninstaller logic and registry backup procedures.
- Building CI/CD pipelines that leverage GitHub Artifact Attestations for transparent binary releases.
Known limitations
This skill focuses on GitHub's Acceptable Use Policies and Windows/Unix system hygiene. It does not replace a full static analysis (SAST) tool or dynamic security auditing.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Security.
Focuses on GitHub policies and Windows/Unix system hygiene. Does not replace professional SAST tools or dynamic security auditing. Requires user to have a Git environment for staging checks.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 3 days ago
- Software
Node.js
Git