October was a big month for us here at Port! As we advance into the era of agentic engineering, we're working harder than ever to bring you features and tools that let you do more.
This month's changelog includes Port AI, BYO LLM, support for multiple organizations, new integrations, support team security improvements, and much more. Read on to see everything new you can do!
Big
Work with Port AI
Port AI is now in open beta! Our interactive chat assistant helps you explore Port’s capabilities, understand your software catalog, and gain actionable insights about your engineering ecosystem through natural conversation. Ask questions like, “Show me all my microservices,” or, “Which repositories haven’t been deployed recently?” and get intelligent answers based on your real catalog data — tailored to your unique blueprint structure and relationships.
If you already have a Port account, you can fill out this form to get access. Please contact our Support team if you have any issues.
Bring Your Own LLM to Port AI
Take full control of Port AI by connecting your own LLM providers, so you can optimize for privacy, cost, and compliance. Choose specific providers and models for different AI interactions — from AI Assistants to AI Agents — and apply fine-grained governance over how your data is processed. Set organization-wide defaults or override them per request for complete flexibility.
Port currently supports Azure OpenAI, AWS Bedrock, Anthropic and OpenAI providers.
Be a member of multiple Port organizations
It's common for teams to build Port in dev, staging, and prod environments.
Today, you can be a member of multiple organizations at the same time with one email address, and switch between them seamlessly. No more juggling accounts.
If you'd like all of your users to automatically join a specific organization, you can enable "Automatic user access" (aka JIT provisioning). When logging in via SSO, your users will automatically join the specified organization.
Don’t see the dropdown? Contact our Support team to migrate your organizations.
Build truly accessible developer experiences
Port now meets WCAG 2.2 (Level A & AA), Section 508, and EN 301 549 standards, making our platform more accessible to developers of all abilities. With enhanced keyboard navigation, screen reader compatibility, improved color contrast, and detailed ARIA labeling, every team member can contribute effectively to your software catalog. For example, the new data model cards view offers a more accessible alternative to the graph view.
Medium
Okta integration
Bring your organization’s users and groups into your catalog by connecting Okta. Start in Data Sources, add Okta, and watch your identity data sync live.
Checkmarx One integration
Monitor code security vulnerabilities directly in Port with our comprehensive Checkmarx One integration. Track security findings, vulnerability trends, and code quality metrics from Checkmarx One scans, alongside your services and repositories.
ArmorCode Integration
Centralize your application security posture with the new ArmorCode integration. Track vulnerabilities, risk scores, and remediation priorities from ArmorCode’s 250+ security tools alongside your services and infrastructure. Monitor trends across your entire stack, automate syncs to keep findings up to date, and enable self-service remediation workflows powered by ArmorCode’s AI-driven risk correlation to focus on what matters most.
Conditionally show or hide action steps
When building self-service actions with multiple steps, you may want to show or hide steps based on certain criteria. For example, if a user chose a certain entity in step 1, you can choose to show step 3 and hide step 2.
Steps can be shown or hidden using a “visible” parameter in the steps array. It accepts JQ.
Small
Remote MCP improvements
Port’s remote MCP now offers a new
count_entities
tool to answer questions like, “How many open pull-requests my team have?” or, “How many failed deployments did the frontend service have in the past week?” We also optimized the tool's context so you can work faster and for longer periods without reaching the context limit and over-utilizing LLMs.Integration audit logs available via API
Use an integration’s audit log to track mapping changes, better troubleshoot issues, and gain more insight into how your integrations work.
Integration enhancements
- Azure DevOps:We now support seven new kinds of resources to help you track your Azure DevOps integration end to end. This include builds, test run, pipeline stage and more.
- GitLab:Releases and tags are now available in the GitLab integration, providing real-time visibility into release management and version tagging across your repositories.
✨ AI Guides
Streamline IaC with AI
Creating infrastructure is tedious; manually writing Terraform for every S3 bucket, RDS instance, or cloud resource slows teams down and risks inconsistency. Port + AI coding agents generate Terraform files automatically, open PRs against your IaC repository, and surface changes for review. Every resource stays governed, compliant, and aligned with your standards, so you can work quickly and securely.
Auto-remediate insecure cloud resources with AI
Security violations sometimes slip through — unencrypted storage, public S3 buckets, missing deletion protection — and manual checks are slow and error-prone, leaving your infrastructure exposed. Port + Claude Code automatically detect insecure cloud resources and generate Terraform patches to fix violations. Enforce security policies in the process of creation and remediate misconfigurations automatically — no manual reviews needed.
Improve specifications with Port AI
Coding agents fail when they lack context. Vague Jira tickets lead to failed runs, wasted time, and incomplete results. Learn how to leverage Port to automatically improve task specifications before they reach your coding agents. Port AI analyzes each issue, fills in missing details, and validates that requirements are clear and actionable. Every ticket becomes well-defined and ready for automation, ensuring coding agents deliver better results, faster.
Manage Port’s support team's access to your environment
You can now manage when Port's support team is able to access your organization's environment.
Port can create dedicated support users in your organization with functionally-appropriate permission levels (
read-only
or admin
), and you can enable or disable Port support access via opt-out controls in your organization's settings. Additionally, you can temporarily grant Port's support team access to your environment: for one day, one week, one month, or indefinitely. Critical incident response from Port's support team is always available, regardless of your organization's opt-out settings.Organizations are opted-in to this feature by default, but these settings can be changed at any time from the Organization settings section of the Builder page. More details are available in Port’s documentation.