This release of Qoder CN introduces a new unit test agent for the AI Programmer, improves code generation from images, and adds one-click reporting for errors. The Enterprise Edition now allows you to configure pre-filters, post-filters, and your own model services.
Qoder CN gets better every day.
The unit test agent is a specialized capability of the AI Programmer that can generate unit test files in batches for code changes, a single code file, or multiple files. You provide the content to be tested and your requirements, and the AI Programmer automatically generates a test plan, creates test cases, compiles and runs the tests, and performs automatic repairs based on error messages. This feature significantly increases test case coverage and quality, reducing the effort of writing unit tests. For more details, see File-level unit test generation (UnitTest).
If you encounter an exception or error while using Qoder CN, or when troubleshooting an issue with the Qoder CN support team, you can click "Report an Issue", describe the problem, and submit it. A feedback code is automatically generated upon submission. You can send this code directly to the Qoder CN support team.
The Standard and Exclusive plans of the Enterprise Edition now allow you to configure your own model services. Organization administrators can configure these services in the Qoder CN Enterprise Edition console, which provides greater flexibility for using the inference service.
In the left navigation pane of the console, choose Policy configuration > Inference service configuration. The page provides configuration areas for the Intelligent Q&A scenario and the AI Programmer scenario. Turn on the Enable model service configuration toggle, enter the Model URL, Model name, and API key, and then click Test connection to verify connectivity.
For intelligent Q&A and intelligent completion scenarios, organization administrators can configure post-filters using regular expressions or custom scripts. Once enabled, these filters are automatically triggered when a developer uses the intelligent Q&A or intelligent completion features in the Qoder CN plugin. This initiates a message push, helping you meet corporate requirements for an information audit.
On the Policy configuration > Filter configuration page, select the Intelligent Q&A tab. You can configure both the Intelligent Q&A pre-filter and the Intelligent Q&A post-filter. After enabling the filter, select a filter type (such as Regular expression), and set the action (Block when rule is matched or Do not process when rule is matched). In the rules table, add a rule by entering a Rule name and a Regular expression. You can enable Message push as needed. When finished, click Save configuration.
1.AI Programmer adds unit test agent
The unit test agent is a specialized capability of the AI Programmer that can generate unit test files in batches for code changes, a single code file, or multiple files. You provide the content to be tested and your requirements, and the AI Programmer automatically generates a test plan, creates test cases, compiles and runs the tests, and performs automatic repairs based on error messages. This feature significantly increases test case coverage and quality, reducing the effort of writing unit tests. For more details, see File-level unit test generation (UnitTest).
2.One-click reporting for Qoder CN plugin errors
If you encounter an exception or error while using Qoder CN, or when troubleshooting an issue with the Qoder CN support team, you can click "Report an Issue", describe the problem, and submit it. A feedback code is automatically generated upon submission. You can send this code directly to the Qoder CN support team.
3.Enterprise Edition supports custom model service configuration
The Standard and Exclusive plans of the Enterprise Edition now allow you to configure your own model services. Organization administrators can configure these services in the Qoder CN Enterprise Edition console, which provides greater flexibility for using the inference service.
In the left navigation pane of the console, choose Policy configuration > Inference service configuration. The page provides configuration areas for the Intelligent Q&A scenario and the AI Programmer scenario. Turn on the Enable model service configuration toggle, enter the Model URL, Model name, and API key, and then click Test connection to verify connectivity.
4.Enterprise Edition supports pre-filter and post-filter configuration
For intelligent Q&A and intelligent completion scenarios, organization administrators can configure post-filters using regular expressions or custom scripts. Once enabled, these filters are automatically triggered when a developer uses the intelligent Q&A or intelligent completion features in the Qoder CN plugin. This initiates a message push, helping you meet corporate requirements for an information audit.
On the Policy configuration > Filter configuration page, select the Intelligent Q&A tab. You can configure both the Intelligent Q&A pre-filter and the Intelligent Q&A post-filter. After enabling the filter, select a filter type (such as Regular expression), and set the action (Block when rule is matched or Do not process when rule is matched). In the rules table, add a rule by entering a Rule name and a Regular expression. You can enable Message push as needed. When finished, click Save configuration.
5.Inline completion enhancements
- Enhanced SQL generation for code completion.
- Improved inline completion performance.
6.Intelligent Q&A enhancements
- Improved the AI Programmer's ability to recall conversation history.
- Streamlined code explanation responses.
- The AI Programmer now applies snippet-level edits to large files faster.
- Improved code generation from images in both intelligent Q&A and the AI Programmer.
7.Other improvements and bug fixes
- Improvement: You can now sign in to the Enterprise Edition with a personal access token in JetBrains IDEs.
- Improvement: Mermaid diagrams are now rendered in a collapsed state by default.
- Improvement: Optimized the file path selection when creating new files with the AI Programmer.
- Improvement: The AI Programmer now supports generating code changes within a project from an image.
- Fixed: An incorrect file path display when the unit test agent generated a test plan.
- Fixed: An issue where Q&A with images would fail if queried before the image finished processing.
- Fixed: An intermittent issue that prevented content from being copied during unit testing.
- Fixed: An intermittent error when opening the Diff View in the workspace.