Fix frontend build due to typescript issues

Another release with a new set of features and fixes!

Tests 🕵️

• You can now define tests with multiple interactions (user inputs and expected agent responses). Tests will fail fast comparing each agent response to the expected one.
• Tests run in the background now, no longer blocking UI, interrupting tests or presenting some inconsistent behaviors when running and user closes tabs, tab deactivates, etc.
• When running tests, proper authentication is requested if the agent uses any tool requiring it (like Jira or some MCP servers)

Misc 🔢

• Now browser copilot expands agents lists by default.
• Change browser copilot icon to better align to the general style.
• Fix visualization issue with sidebar sections when sidebar is collapsed

This release includes a new set of features to enjoy! 🎉

Misc 🔢

• You can now add users as editors in a group. Editors can edit public agents and prompts in a team, but can't manage team members or see team metrics.
• Add DISABLE_PUBLISH_GLOBAL environment variable that allows to disable any member of the global team to publish agents to it. This is handy when you need more control over the agents that are published to everyone that has access to your Tero instance.
• Remove no longer necessary INTERNAL_EVALUATOR_TEMPERATURE environment variable
• Fix some authentication issues when Tero tab is pinned for a long period of time.

Chat 💬

• This version fixes an issue that prevented users that didn't configure an agent, to use it if the agent uses using Jira or MCP tool.
• Improved thought process visualization.

Browser Copilot 🤖

• Fix token renewal which caused Tero instances to stop working in the Copilot unless a manual refresh was done.

This release includes several new features! 💯

Tests 🕵️

• You can no customize the evaluator used in agents tests. You can set the prompt, the model and model temperature or reasoning level. This allows you to focus your evaluator on the specific needs of your agent or in the specific needs of a particular test.
• Now auto generated test cases names are meaningful, no longer Test Case #X, and based on the test case messages.
• You can now also review past executions to detect deviations in agent behavior.
• You can clone tests to speed up test suite creation
• You can also compare side-by-side a given test result with the expected result.
• We have improved default evaluator to avoid getting some false positives.

Chat 💬

• Thought process now appears immediately after you send a message, and is persistent, which allows you to review past chats through processes and compare with new ones to detect deviations in agent behavior.
• Prompt variables dialog is now bigger by default, which helps when variables input values are long.

Agent Edition ✏️

• Now models are categorized by vendor which eases finding them in the drop downs.

Misc 🔢

• This version includes an authentication fix regarding tokens refresh that generated some inconsistent behavior with session handling.
• When uploading some csv files now there is a fallback mechanism when encoding detection fails.
• Now Tero shows a warning instead of failing when an unexpected variable is set in .env file.

Browser Copilot 🤖

• You can now remove a Tero instance and individually refresh each Tero instance agents

This version fixes import of agents that contains some special characters (like ñ) in some tool file in windows.

Thanks to @ldelreguero for reporting this issue.

Now you can collapse and expand agents and chats sections in sidebar!

Additionally, "New Chat" button includes a tooltip that clarifies it's purpose.

Thanks to @ldelreguero for these contributions!

This release fixes doc tool embeddings usage when using Azure (an issue introduced in version 1.0.2) and includes a small improvement in UI when importing files and changes default team name to Global por consistency.

Thanks to @ldelreguero to report some of these issues.

This version allows to import existing OSS agents in a Tero instance that doesn't configure advanced file processing, and fixes handling with some special characters in agents file names.

Thanks to @ldelreguero for reporting this!

This version allows to use Tero only configuring OpenAI API keys.

In previous versions, even though the documentation said that only OpenAI API keys were required, to use OpenAI you still needed to configure some Azure OpenAI settings. Now it is not necessary.

Thanks to @ldelreguero to report this!

Fix docker-compose references to common module for proper startup when running docker-compose up.

Thanks to @germainvenzal and @ldelreguero to report and help fix this.