Release Notes - August 2026

Expanded Tool Coverage for Trust Agent AI

What's New:
Trust Agent AI, which uses artificial intelligence (AI) to give visibility into how your developers use AI coding tools, now captures commit activity from three additional tools: Cursor, Copilot CLI, and Devin (formerly Windsurf). These tools submit commit analysis through the same integration already used for Claude Code and Codex CLI, so their activity appears in commit analysis reports automatically. Where a tool reports which model was used, that usage now also flows into large language model (LLM) policy reporting.

Key Benefits:

• Commit analysis reports reflect a fuller picture of AI tool usage, including tools that weren't previously visible
• No additional setup is required to start seeing data from Cursor, Copilot CLI, or Devin
• Model usage from these tools is included in LLM policy reporting alongside existing tools

How to Access:
Available now for customers using the local agent on insightsd v0.2.2 or later.

Topic Descriptions in Quest Creation

What's New:
When selecting objectives while building a Quest, a Curriculum Preview panel now appears alongside the topic list. Each selected topic can be expanded to show a description of what it covers, so admins no longer have to guess based on the topic name alone.

Key Benefits:

• Admins can see what a topic covers before adding it to a Quest
• Descriptions can be expanded or collapsed individually or all at once
• Topics can be removed directly from the preview panel

How to Access:
Available now for all customers using Quests. Open the objective selection step when creating or editing a Quest to see the Curriculum Preview panel.

Automatic Scheduled Vulnerability Data Sync

What's New:
API-based vulnerability data importers for Trust Agent now support continuous background sync, so admins no longer need to manually re-run imports. When setting up a vulnerability data connection, admins can choose a one-time import or a continuous monthly schedule, and can name each connection to keep multiple sources organized. Imported data aggregates across all sync runs, and findings update automatically as they're resolved or closed in the source security tool.

Key Benefits:

• Vulnerability data stays current automatically, without manual re-imports
• Multiple data connections can be named and managed independently
• Findings reflect resolved or closed status from the source tool automatically

How to Access:
Available now for all customers. Go to Administration > Trust Agent configuration > Vulnerability signal > Import to set up or manage scheduled imports.

New AI Challenges Across Six Languages

What's New:
68 new AI Challenges have been added to Quest topics across six languages, including pseudocode and mobile languages: C Basic, C++ Basic, Java Android SDK, Kotlin Android SDK, Pseudocode Web, and Swift iOS SDK. Explore now includes roughly 1,127 AI Challenges in total.

Key Benefits:

• Learners get expanded hands-on practice across more languages and platforms
• New challenges are automatically available in Quests, Learn, and Explore
• The "New content" filter makes it easy to find recently added challenges

How to Access:
Available now for all customers. Use the "New content" filter at the top right of the Learn and Explore pages to find the new challenges.

New Interactive Walkthrough for Threat Modeling with AI

What's New:
A new interactive walkthrough mission has been added to the Threat Modeling with AI topic, available in Quests and Learn. It's designed to help learners apply critical thinking to AI-generated output by choosing the right clarifying questions to ask.

Key Benefits:

• Learners practice evaluating AI output rather than accepting it at face value
• The walkthrough reinforces critical thinking skills specific to working with AI tools
• Available alongside existing Threat Modeling with AI content in Quests and Learn

How to Access:
Available now. Find it within the Threat Modeling with AI topic in Quests or Learn.

Cybermon: Event Countdown and Achievements Page

What's New:
The Cybermon event page went live with a homepage banner and a live countdown to this year's event. Learners also now have a dedicated achievements page where they can view all past Cybermon badges, grouped by year, see which ones they've earned, and share earned badges directly to LinkedIn.

Key Benefits:

• Learners get a clear, visible countdown to the Cybermon event from their homepage
• Past achievements are organized in one place and viewable by year
• Earned badges can be shared directly to LinkedIn from the achievements page

How to Access:
Available now for all learners. The countdown banner appears on the Learner Homepage, and achievements are available in the homepage achievements section.

New Interactive Walkthroughs for Docker

What's New:
Eight new interactive walkthrough missions have been added to the Docker content within the container and DevOps offering, bringing most Docker topics up to a full range of interactive activity types.

Key Benefits:

• Learners get more hands-on practice across Docker topics
• Content variety brings Docker in line with other container and DevOps topics
• Addresses customer feedback comparing SCW's lab offering to competitors

How to Access:
Available now for all customers with access to the Docker container/DevOps content.

Citizen AI: Automating with AI Content Complete

What's New:
The initial Citizen AI content offering is now complete, with four interactive walkthroughs in each of the four topics in the Automating with AI bundle: From Chat to First Automation, When Things Break, What Can Go Wrong and Limiting the Risks, and Identity, Credentials, and Access.

Key Benefits:

• Customers now have a complete set of walkthroughs covering practical AI use for non-technical roles
• Content is ready to support proof-of-value engagements for the Citizen AI bundle
• Scenarios are task-based, making them applicable across a range of roles

How to Access:
Available now for customers with access to the Citizen AI bundle.

Learning Budget for Adaptive Learning with Vulnerabilities

What's New:
Admins can now set a training time budget for Adaptive Learning with Vulnerabilities, limiting the total number of training topics assigned to each developer per 90-day cycle. When a budget is set, the platform ranks each developer's repositories by recent commit activity and assigns training starting from the most active repository, stopping once the next repository would exceed the budget. Without a budget set, behavior is unchanged.

Key Benefits:

• Admins can roll out Adaptive Learning without overwhelming developers who commit across many repositories
• Training is prioritized based on each developer's most active repositories
• Existing behavior is preserved for admins who don't set a budget

How to Access:
Available now for all customers using Adaptive Learning with Vulnerabilities.

Adaptive Learning with Trust Agent Vulnerabilities