Average Ratings 0 Ratings

Total
ease
features
design
support

No User Reviews. Be the first to provide a review:

Write a Review

Average Ratings 0 Ratings

Total
ease
features
design
support

No User Reviews. Be the first to provide a review:

Write a Review

Description

Codename MDASH represents an advanced code scanning tool integrated within Microsoft Defender, leveraging a multi-modal AI framework to uncover, verify, and address vulnerabilities with a level of insight that surpasses conventional static analysis methods. This system enhances the Defender CLI by introducing a multistage process where specialized agents work collaboratively through four distinct phases. Initially, the preparation stage organizes ranked files based on risk, utilizing call-graph analysis and evaluating code complexity to identify functions with the highest likelihood of containing vulnerabilities. The scanning phase then transmits this prioritized code to over 100 specialized agents, including those focused on injection flaws, memory safety issues, and authentication bypasses, ensuring each agent targets a specific category of vulnerability. Following this, the validation phase employs taint analysis, type resolution through Language Server Protocol servers, and a debate among multi-model agents to improve confidence levels and minimize false positives. Finally, the deduplication step merges overlapping findings into a comprehensive set of unique and actionable results, enhancing the overall efficiency of vulnerability management processes. This innovative approach signifies a new era in threat detection and remediation within software development.

Description

Wardstone functions as a security API for language models, acting as a safeguard between applications and various language model providers by evaluating inputs and outputs for potential threats in four distinct categories during a single request: prompt attacks, content violations, data leaks, and suspicious links. It is adept at identifying jailbreaks, prompt injections, harmful content including hate speech, violence, and self-harm, as well as personally identifiable information like Social Security numbers, credit card details, email addresses, and phone numbers, in addition to detecting dubious URLs. Each response generated provides a detailed risk assessment for each category, achieving this within a swift response time of under 30 milliseconds. Compatible with any LLM provider, it is accessible via a REST API and offers SDKs for multiple programming languages, including TypeScript, Python, Go, Ruby, PHP, Java, and C#. A complimentary tier allows up to 10,000 calls per month without the need for a credit card, and it also features a web-based playground for users to experiment and test functionalities. Users can easily integrate this tool into their existing systems, enhancing the security of their language model interactions.

API Access

Has API

API Access

Has API

Screenshots View All

Screenshots View All

No images available

Integrations

MAI-Cyber-1-Flash

Integrations

MAI-Cyber-1-Flash

Pricing Details

No price information available.
Free Trial
Free Version

Pricing Details

$0/month
Free Trial
Free Version

Deployment

Web-Based
On-Premises
iPhone App
iPad App
Android App
Windows
Mac
Linux
Chromebook

Deployment

Web-Based
On-Premises
iPhone App
iPad App
Android App
Windows
Mac
Linux
Chromebook

Customer Support

Business Hours
Live Rep (24/7)
Online Support

Customer Support

Business Hours
Live Rep (24/7)
Online Support

Types of Training

Training Docs
Webinars
Live Training (Online)
In Person

Types of Training

Training Docs
Webinars
Live Training (Online)
In Person

Vendor Details

Company Name

Microsoft

Founded

1975

Country

United States

Website

learn.microsoft.com/en-us/security-exposure-management/ai-code-security-overview

Vendor Details

Company Name

JRL Software LTD

Founded

2024

Website

wardstone.ai

Product Features

Product Features

Alternatives

Claude Security Reviews

Claude Security

Anthropic

Alternatives

CodeMender Reviews

CodeMender

Google DeepMind