1. 1AI Asset Discovery
  2. 2Static Analysis
  3. 3Behavioral Testing
  4. 4Red Teaming
  5. 5Risk Analysis & Reporting
  6. 6Integrations

Integrations

Plug into tools you already use.

In practice

A platform team's ServiceNow records now update with every rescan of its support agent, so governance sees the agent that is actually running and nobody types it in by hand.

  • No more re-typing AI assets into ServiceNow.
  • Governance records that match what is running.
  • Findings land in the tools engineers already check.

Set up the integration you need

Pick an integration to see its documented setup and what it adds.

Integrations / permits-assistantSARIFCycloneDXSPDX
ServiceNow AI Control TowerNuGuard platformClaude CodeClaude Code pluginClaude DesktopMCP serverCursor, Windsurf and ClineMCP serverCI pipelineOpen-source CLIGitHub Code ScanningOpen-source CLICycloneDX and SPDXOpen-source CLI

Setup

  1. In Settings > Integrations > ServiceNow, enter your instance's base URL, a service account's user ID and password, and an OAuth application's client ID and secret.
  2. Select Save Credentials, then Test. Turn on Enable once the test passes.
  3. Open an application, choose a completed scan, and select Export to ServiceNow in its AIBOM view.
NuGuard platform

ServiceNow AI Control Tower

What it adds

  • An AI System record for the application in AI Control Tower
  • Model, prompt and dataset records linked to that AI System
  • Created, found and error counts after each export
  • Re-exporting a new scan can update records already in ServiceNow

Simplified sample view. Commands and config follow the NuGuard docs.

How scan results reach ServiceNow

  1. Connect ServiceNow onceService account and OAuth app, in Settings
  2. Run a scanAn AIBOM scan through Asset Discovery
  3. Export the scanFrom the AIBOM view, or Reports > Integrations

Click a step to play it again.

Where NuGuard plugs in

ServiceNow AI Control Tower

AI assets from a scan, exported as linked records

Claude Code plugin

Slash commands, two skills and a security-auditor agent

MCP-compatible editors

Claude Desktop, Cursor, Windsurf and Cline, with seven tools

CI pipelines

Gate builds at a severity and keep the reports

GitHub Code Scanning

SARIF from scans, static analysis and red teaming

CycloneDX and SPDX

AI-SBOM exports, including model cards

See it in the product

Integrations FAQs

What are Integrations?

Integrations put NuGuard where your team already works: a ServiceNow AI Control Tower export on the NuGuard platform, plus coding agents, CI pipelines and standard export formats. Findings that live only inside NuGuard don't get fixed, and re-typing AI assets and findings into ServiceNow by hand doesn't scale past the first scan.

Is the ServiceNow integration a one-time setup?

The connection is. An Editor or Admin adds the credentials in Settings > Integrations > ServiceNow, tests the connection and enables it. Exports are per scan: choose a completed scan and select Export to ServiceNow, then export again after a rescan.

What does the ServiceNow export send?

An AI System record for the application, plus the models, prompts and datasets the scan found, linked to it in AI Control Tower. Each export reports how many records it created, found already in ServiceNow, or failed on, and re-exporting can update records that already exist.

Can I run NuGuard from Claude Code or another coding agent?

Yes. The coding-agent integrations run the open-source NuGuard engine, the same one the NuGuard platform calls. The Claude Code plugin installs from the plugin marketplace and adds the /nuguard-scan, /nuguard-redteam, /nuguard-config and /nuguard-init commands, the ai-security-review and sbom-analysis skills, and a security-auditor agent. Claude Desktop, Cursor, Windsurf, Cline and other MCP-compatible editors connect to NuGuard as an MCP server with seven tools, and Smithery can install that server for you.

How do I add NuGuard to a CI pipeline?

The NuGuard platform and the open-source CLI share one engine, so a pipeline can run the CLI directly: install it with pip install nuguard, then run nuguard scan --source . --output-dir nuguard-reports --fail-on high. The scan exits with code 1 when a finding reaches the threshold, 2 on any critical finding and 3 on a scan error, and it writes report.md, findings.json and findings.sarif to nuguard-reports, so you can keep that folder as a build artifact and upload findings.sarif to GitHub Code Scanning with the github/codeql-action/upload-sarif action.

Can I export the AI-SBOM to CycloneDX or SPDX?

Yes. On the CLI, nuguard sbom plugin run exports an AI-SBOM to CycloneDX 1.6, to CycloneDX 1.6 with AI-specific extensions such as model cards, or to SPDX 3.0.1 JSON-LD. The same command has a sarif_export plugin for GitHub Code Scanning and a markdown_export plugin for a readable SBOM report.

Previous in the NuGuard pipeline

Previous in the pipeline: step 5 of 6

Risk Analysis & Reporting

Get a ranked fix list and a release gate

Related reading

Put NuGuard results where your team already works.

Try NuGuard Free
  • 1. AI Asset Discovery
  • 2. Static Analysis
  • 3. Behavioral Testing
  • 4. Red Teaming
  • 5. Risk Analysis & Reporting
  • 6. Integrations