Skip to main content

Release 5.0 2026.08.11.1

· 3 min read

New Features 🚀​

AI Services​

Serenity* Star keeps expanding the AI Services available through its API, so your AI solutions can rely on the platform for more than agents alone.

Anonymizer​

The new Anonymizer service detects and obfuscates sensitive data in any text. You decide which categories to detect (person, address, email, phone number, dates and more), the confidence threshold for each one and the replacement label used to build its placeholder. The service returns the obfuscated text together with every entity it detected, and it runs in a secure local environment, so you can redact personal data before sending it to an external LLM.

alt text

Inference Playground​

The Inference service now comes with a playground, so you can test it straight from the web app. Choose a protocol, pick an example such as streaming or vision, and get instant responses without writing a line of code.

alt text

Serenity* Apps​

SEO Analysis​

The SEO Analysis app gives you an at-a-glance view of your site's exposure, together with insights into your SEO and GEO positioning. It surfaces actionable items so you can improve your position quickly and easily, without relying on an external agency.

alt text

Meeting Minutes​

The Meeting Minutes app turns a meeting transcript into structured minutes, covering what was discussed, the decisions taken and the action items that came out of them. Share the result with your team so nothing agreed in the meeting is lost.

alt text

Enhancements ✨​

Execution Naming​

Agent instances are now named automatically, giving you the context of what happened in each conversation at a glance and making executions much easier to track on the grid.

alt text

Onboarding​

Company onboarding now lets you add relevant files and choose which skills are available in the AI Chat, all from a single simplified view.

alt text

alt text

Agent Reasoning Detail​

The new reasoning detail configuration lets you decide how much of your agent's reasoning is exposed, from short summaries to full explanations.

Only available on models that support reasoning.

New Agent Templates​

We have reworked how agent templates are offered, so you can start your AI journey with more powerful agents that take advantage of the full set of Serenity* Star features.

Knowledge Performance Improvement​

The agent Knowledge section has been optimised for large numbers of files. No more laggy screens, just a smoother working experience.

Bug Fixes & Minor Improvements 🛠️​

  • Fixed the autosize behaviour of the JSON output format field when an agent is loaded.
  • Improved validation in the OCR service.
  • Improved validation when uploading images for the chat widget icon, along with an informative caption describing the recommended image format.
  • Fixed an issue where the permissions selector did not load users on Proxy agents.
  • Agent permission policies are now applied to the OpenAI-compatible proxy endpoints.
  • Improved dataset schema validation when replacing and appending tables.
  • Fixed an issue when copying the Serenity* AI Chat channel on Chat agents.
  • An invalid request body now returns a bad request error from the API instead of an internal server error.
  • Added a new endpoint to retrieve the details of a conversation instance, with optional includes.
  • Improved the feature validator when purchasing an agent from the store.
  • Fixed an error when retrieving an agent version through the API when its skills had no AI models configured.