Skip to content
KitploitKITPLOIT
ToolsExploitsBlog
Log in
Submit
ToolsExploitsBlog
Submit

Hacking, PenTest, and Cybersecurity Tools for Your Security Arsenal!

Kitploit is a directory of hacking, cybersecurity, and pentesting tools. Discover the latest project updates to find vulnerabilities, analyze systems, automate testing, and strengthen your security.

FeedsContactPrivacy© 2026 Kitploit

Tool Directory

Categories

View all categories
Loading categories
GuardModel — GitHub Action that scans ML model files for malicious code and security vulnerabilities | Kitploit
Tools/GitHubGitHub/shivang0/guardmodel
Static AnalysisVulnerability ScannersCode AnalysisMalware AnalysisDevSecOpsSecret DetectionSupply Chain SecurityAI Security
GitHubshivang0/guardmodel

GuardModel

GitHub Action that scans ML model files for malicious code and security vulnerabilities

View Repository
25111 months agoNot yet reviewed
Share

GuardModel

Block malicious AI models in your pipeline with one line of YAML.

GuardModel is a GitHub Action that automatically scans ML model files for malicious code, vulnerabilities, and security risks in CI/CD pipelines. It blocks dangerous models before they reach production by detecting pickle deserialization attacks, embedded malware, and known CVEs.

Quick Start

Add to your workflow:

name: GuardModel Security Scan
on: [push, pull_request]

jobs:
  scan:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: guardmodel/scan@v1

That's it! GuardModel will scan all model files and fail the check if threats are detected.

Features

  • Automatic Detection - Scans all model files on push/PR
  • Multiple Formats - Supports pickle, PyTorch, Keras, ONNX, SafeTensors
  • 70+ Detection Rules - Code execution, reverse shells, file system access, network activity
  • GitHub Integration - SARIF output for Security tab, PR comments, JSON reports
  • Configurable - Allowlists, severity thresholds, custom rules

Supported Formats

ExtensionFormatScannerRisk Level
.pkl, .picklePython PicklePickle AgentCritical
.pt, .pthPyTorchPickle AgentCritical
.binPyTorch/TransformersPickle AgentCritical
.h5, .hdf5HDF5/KerasKeras AgentHigh
.kerasKeras v3Keras AgentHigh
.onnxONNXONNX AgentMedium
.safetensorsSafeTensorsSafeTensors AgentLow

Configuration

Create .guardmodel.yml in your repository:

version: 1

# Directories to scan
include:
  - models/
  - weights/

# Directories to ignore
exclude:
  - tests/fixtures/

# Severity threshold to fail CI
fail_on: high  # critical, high, medium, low, none

# Allowlist known-safe models by SHA256
allowlist:
  - sha256: "abc123..."
    reason: "Verified by security team"

# Maximum file sizes
max_file_size: 5GB
max_total_size: 20GB

Action Inputs

InputDescriptionDefault
pathDirectory to scan.
configConfig file path.guardmodel.yml
fail-onMinimum severity to failhigh
output-sarifGenerate SARIF outputtrue
output-jsonGenerate JSON outputtrue
comment-on-prPost PR commenttrue
max-file-sizeMaximum file size5GB

Action Outputs

OutputDescription
statusScan status (passed, failed, error)
findings-countTotal findings
critical-countCritical severity findings
high-countHigh severity findings
sarif-filePath to SARIF output
json-filePath to JSON output
scan-durationDuration in milliseconds

Threat Detection

GuardModel detects:

  • Code Execution - os.system, subprocess, eval, exec
  • Reverse Shells - Socket connections with shell execution
  • Network Activity - HTTP requests, downloads, data exfiltration
  • File System - File deletion, permission changes
  • Dangerous Imports - ctypes, dynamic imports
  • Obfuscation - Base64, marshal, nested pickles
  • Known Malware - Hash and signature matching
  • CVE Vulnerabilities - Known security issues

Example Output

PR Comment

GuardModel Security Scan

Status: FAILED - 2 threats detected

FileSeverityThreatDetails
models/model.pklCriticalCODE_EXECUTIONos.system call detected
models/utils.ptHighNETWORKsocket.socket creation

SARIF Integration

Findings appear in the GitHub Security tab with:

  • Rule descriptions
  • File locations
  • Remediation guidance

Development

# Install dependencies
npm install
pip install -r requirements.txt

# Build
npm run build

# Test
npm test
pytest tests/

# Package for release
npm run package

Architecture

guardmodel/
├── action.yml              # GitHub Action definition
├── src/                    # TypeScript source
│   ├── index.ts           # Entry point
│   ├── orchestrator.ts    # Scan coordination
│   ├── config.ts          # Configuration
│   ├── walker.ts          # File discovery
│   └── reporters/         # Output formatters
├── python/
│   ├── agents/            # Scanner agents
│   │   ├── pickle_agent.py
│   │   ├── keras_agent.py
│   │   ├── onnx_agent.py
│   │   └── safetensors_agent.py
│   └── rules/             # Detection rules
└── tests/                 # Test suite

Security

GuardModel:

  • Scans locally - no files uploaded to external servers
  • No telemetry without opt-in
  • Runs in isolated GitHub runner environment
  • Dependencies pinned with hash verification

License

MIT

Contributing

Contributions welcome! Please see CONTRIBUTING.md.

Links

  • Documentation
  • Rule Reference
  • GitHub Marketplace
Download Tool

Most Popular

View all →

Discover the most used tools by our community.

  • 1

    OsintHunter

    letox4/osinthunter

    🎯 The ultimate OSINT toolkit: investigation graph, confidence scoring, alias families, map, timeline and reports. 49 modules, 400+ sites, no API keys required.

    837
  • 2

    ghostlock-app

    yukonga/ghostlock-app

    GhostLock One-Tap Execution App (CVE-2026-43499)

    536
  • 3

    ARTEX

    baarai/artex

    LLM multi-agent autonomous penetration testing platform with planner/worker agents, asset and exploration graphs, MITM traffic recording, and vulnerability retesting.

    424
  • 4

    redforge

    cybersecurity91/redforge

    Browser-side red team toolkit with 102 offline tools for recon, payload generation, credential attacks, crypto, encoding and engagement reporting.

    355
  • 5

    looking-glass

    anonghosty/looking-glass

    CSIRT platform that ingests Shadowserver reports into MongoDB and replays them as interactive maps, analysis, findings, advisories and executive dashboards.

    294
  • 6

    CyberStrikeAI

    ed1s0nz/cyberstrikeai

    The system of action for AI-native cybersecurity—where intent becomes governed execution, evidence becomes operational memory, and every operation improves the next.

    276

Explore all tools

Browse our collection of tools

View all tools →