NSFOCUS AI Guardrails · emploidai Marketplace
emploidai Marketplace
Add-onsAppletsPlugins
Search tools, teams, and capabilitiesPublish
MarketplacePluginsNSFOCUS AI Guardrails
Plugin
Limited listing

NSFOCUS AI Guardrails

by nsfocus · v0.0.2

NSFOCUS AI Guardrails is a professional content security protection component launched by NSFOCUS based on the fighting AI with AI principle. It filters sensitive content and malicious attacks in inputs and outputs in real time, ensuring AI applications are compliant and reliable, with easy integration.

395 installsUpdated Feb 2, 2026
Publisher information is incomplete

This community listing does not yet include every recommended support, privacy, pricing, and permission disclosure. Review the available package permissions before installing.

Capabilities

Tools

Available inside your emploidai workspace after installation.

Data sources

Available inside your emploidai workspace after installation.

Category

tool

Version

0.0.2nsfocus

Requirements

Maximum memory 256MB

Pricing

Not disclosed by publisher

Security & access

Review before installing

CompatibleRequires emploidai 1.0.0+

Permissions

  • Requires encrypted tool credentials

Dependencies

No additional dependencies

Resources

Privacy policy
emploidai Marketplace

Discover capabilities. Review access. Install inside your workspace.

DocumentationSecuritySupportPrivacyTerms

NSFOCUS AI Guardrails is a professional content security protection component launched by NSFOCUS based on the "using models to govern models" concept. It filters sensitive content and malicious attacks in input and output in real-time, ensuring the compliance and reliability of AI applications, and can be easily integrated.

Operational Steps

  1. Log in to the Guardrails management interface as a user.

  2. Complete asset onboarding. For detailed instructions, please refer to the "Asset Onboarding Steps". Record the AssetID for each asset, as it will be required for subsequent tool calls.

  3. Click the user avatar in the upper right corner to obtain your AK and SK.

  4. Configure the AK/SK into the plugin.

Input/Output Text Detection

Detects textual content in both input and output directions. For the input direction, it checks content compliance and for semantic-layer attacks. For the output direction, it checks content compliance.

Parameters:

  • Detection Type: select Options: Input Detection / Output Detection.
  • Asset ID: string The asset's AssetID, obtainable from the Asset Management page of NSFOCUS AI Guardrails.
  • Query:string The text content input by the user.

Output:

  • errCode: number Error code. 200 indicates normal.
  • errMsg: string Error message.
  • suggestion: string Judgment result. Values: pass/block/error
  • answer: string Suggested replacement answer provided.
  • categories: array[string] Categories determined for the detected content.

Input/Output File Detection

Detects files and images uploaded in input and output directions. Converts content from files (Word, PDF, PPT, etc.) to text for compliance checking; understands image intent to check image compliance.

Supported File Types: PDF / Word / Excel / PPT / Visio / WPS / txt Supported Image Types: JPG / JPEG / PNG / GIF / BMP / TIFF / TIF

Parameters:

  • Detection Type: select Options: Input Detection / Output Detection.
  • Asset ID: string The asset's AssetID, obtainable from the Asset Management page of NSFOCUS AI Guardrails.
  • File Object: File The file/image input by the user.

Output:

  • errCode: number Error code. 200 indicates normal.
  • errMsg: string Error message.
  • suggestion: string Judgment result. Values: pass/block/error
  • answer: string Suggested replacement answer provided.
  • categories: array[string] Categories determined for the detected content.

Service Access Methods

  1. Scan the QR code to add the WeChat account and contact the product manager to obtain the Guardrails installation package and deployment guidance.
  1. Call the company hotline 400-818-6868 to contact the local representative office for project support.