Release highlights:

• DREAD Risk Scoring: STRIDE GPT now supports DREAD risk scoring, allowing users to assign risk scores to identified threats based on the DREAD model. This feature provides a more comprehensive threat assessment and helps prioritise mitigation efforts.

• Gherkin Test Cases: Users can now generate Gherkin test cases based on the identified threats. This feature helps bridge the gap between threat modelling and testing, ensuring that security considerations are integrated into the testing process.

• UI Enhancements: I've refreshed the user interface making it easier to navigate and interact with the application and its features.

Release highlights:

• Multi-Modal Threat Modelling: STRIDE GPT now supports multi-modal threat modelling using OpenAI's GPT-4o and GPT-4-Turbo models. Users can provide an image of an architecture diagram, flowchart, or other visual representations of their application to enhance the threat modelling process.
• Google AI Integration: I've added support for Gemini 1.5 Pro via the Google AI API. Please note that Gemini doesn't consistently generate JSON output so you may need to retry some requests. In addition, Attack Trees can't be generated using Google AI models because of Google's safety restrictions.
• Refactored Codebase: I've refactored some parts of the codebase to improve maintainability and readability. This should make it easier to add new features and enhancements in future releases.
• Bug Fixes: Minor bug fixes and error handling improvements.

Release highlights:

• Mistral API Integration: Users can now choose to use LLMs provided by Mistral AI to generate threat models, attack trees and mitigation suggestions. This provides an alternative to OpenAI's GPT models, offering greater flexibility and choice for users.

• Refined Prompts: With more people using STRIDE GPT for work, I've updated the threat model prompt templates to encourage the LLMs to generate more comprehensive outputs. Users should now see multiple threats identified within each STRIDE category.

• Public Roadmap: I've created a public roadmap to provide visibility into upcoming features and improvements.

• UI Enhancements: I've made some minor updates to the UI to accommodate the new Mistral API integration and improve the overall user experience.

Release highlights:

• Azure OpenAI Service Integration: Users can now opt to use OpenAI 1106-preview models hosted on the Azure OpenAI Service, in addition to the standard OpenAI API.
• Docker Container Image: To make it easier to deploy STRIDE GPT on public and private clouds, the tool is now available as a Docker container image on Docker Hub.

Release highlights:

• Integration of New GPT Models: The application now supports the latest "gpt-4-1106-preview" and "gpt-3.5-turbo-1106" models, offering advanced capabilities and more accurate responses for threat modelling and attack tree generation.
• Direct OpenAI API Calls: STRIDE GPT now makes direct calls to the OpenAI API in order to take advantage of the recently introduced JSON Mode. This should greatly reduce the reduce the likelihood of syntax errors when generating threat models.
• Refined Attack Tree Generation: The process for generating attack trees has been overhauled to be more reliable, minimising syntax errors when generating Mermaid diagrams and improving the overall quality of the visualisations.
• New Logo and Color Scheme: A refreshed colour scheme and new logo (generated by DALL·E 3).
• Continued Bug Fixes and Performance Improvements: I've made a small number of additional updates to address existing bugs and optimise the application for better performance, ensuring a smoother and more efficient user experience.

Release highlights:

• Threat Mitigations: STRIDE GPT can now suggest potential mitigations for the threats identified in the threat modelling phase. This helps users develop strategies to prevent or minimise the impact of the identified threats.
• Downloadable Output: Users can now download the generated threat model, attack tree, and mitigations as Markdown files directly from the application. This makes it easy to share and document the generated outputs.
• Improved User Interface: I've further refined the user interface to provide a smoother and more intuitive user experience. The application layout has been optimised for better readability and usability.
• Updated GPT Models: STRIDE GPT now supports the latest 0613 versions of the GPT-3.5-turbo and GPT-4 models. These updated models provide improved performance and increased control over the generated output.
• Bug Fixes and Performance Enhancements: I've addressed several bugs and made performance improvements to ensure a more stable and responsive application.

Release highlights:

• Attack Tree Generation: In addition to generating threat models, STRIDE GPT can now generate attack trees for your applications based on the provided details. This helps users better understand potential attack paths for their applications.
• Attack Tree Visualisation: This is an experimental feature that allows users to visualise the generated attack tree directly in the app using Mermaid.js. This provides a more interactive experience within the STRIDE GPT interface.
• GPT-4 Model Support: STRIDE GPT now supports the use of OpenAI's GPT-4 model, provided the user has access to the GPT-4 API. This allows users to leverage the latest advancements in GPT technology to generate more accurate and comprehensive threat models and attack trees.
• Improved Layout and Organisation: I've restructured the app layout to make it easier to navigate and use. Key sections, such as Threat Model and Attack Tree, are now organised into collapsible sections for a cleaner and more intuitive user experience.

Initial release of the application.