astrio-ai/l2m

AI Summary: This issue proposes the removal of a deprecated alias named `remove_reasoning` from the `src/core/models.py` file. The goal is to clean up existing deprecation warnings by eliminating this outdated alias.

AI Summary: This issue proposes enhancements to the way code differences (diffs) are displayed. Key improvements include adding syntax highlighting, offering better color schemes for readability, and implementing a side-by-side comparison view.

AI Summary: This issue aims to improve the performance of the project by optimizing how dependencies are loaded. The core tasks involve implementing more instances of lazy loading, reducing the overall time it takes for the application to start, and identifying and fixing performance bottlenecks through profiling.

AI Summary: This issue proposes adding a caching mechanism for repository structure analysis to improve the startup time for large repositories. The cache should be invalidated whenever file changes are detected.

AI Summary: This issue proposes adding short command aliases to a command-line interface (CLI) tool. The goal is to improve user experience by allowing users to type abbreviated commands, such as '/a' for '/add' or '/c' for '/commit', and also enhance tab completion functionality.

AI Summary: This issue proposes adding a 'dry-run' mode to a modernization tool. This feature would allow users to preview proposed changes, see differences before committing, and estimate associated costs without actually applying any modifications.

AI Summary: This issue proposes enhancing error handling tests by adding more edge cases to `test_exceptions.py`. It also aims to test the retry logic for API failures and scenarios involving graceful degradation.

AI Summary: This issue requests the addition of unit tests for the visualization functions within the `evals/codebleu/visualize.py` module. The tests should cover various result formats and edge cases such as empty results or missing data fields.

AI Summary: This issue involves addressing several TODO comments within the `base_coder.py` file. The tasks include checking the impact of changes on image messages, reviewing token count implications for image messages, and refactoring the `partial_response_function_call` to be pushed into subclasses. These are likely code maintenance and refactoring tasks.

AI Summary: This issue proposes integrating `garak`, an LLM vulnerability scanner, into the L2M system to identify and mitigate security risks. The goal is to prevent the LLM, which handles sensitive data and code, from being exploited through prompt injection or other attacks, ensuring it doesn't leak information or execute dangerous commands.

AI Summary: This issue proposes integrating a Large Language Model (LLM) vulnerability scanner, Garak, into the L2M system to validate Bring Your Own Model (BYOM) LLMs. The goal is to detect potential issues like data leakage, hallucinations, jailbreaks, toxicity, and behavioral regressions before accepting vendor models or when model configurations change, acting as a CI gate.

AI Summary: This issue requests an update to the existing CI workflow. The specific changes and implementation details are currently undefined, but the goal is to deliver a complete `ci.yml` file. This is a placeholder for a more detailed task that will be fleshed out later.

AI Summary: This issue addresses a bug where the CLI and modernization pipeline fail on Windows due to incorrect path handling. The fix involves ensuring cross-platform compatibility by using `pathlib.Path` and testing on Windows environments, specifically checking absolute and relative path resolution.

AI Summary: This issue proposes adding an AGENTS.md file to the root directory of the L2M project. This file will serve as a README for AI coding agents, providing them with context and instructions for working on the project. The goal is to create a standardized location for agent-specific information.

AI Summary: This issue proposes adding tab completion for file paths in the CLI to improve user experience. Currently, users must manually type file paths, which can be cumbersome. Implementing this feature would make the CLI more user-friendly and efficient.

AI Summary: This issue requests the addition of unit tests for interactive CLI commands within the `l2m/tests/` directory. The goal is to ensure the functionality of these commands and that all tests pass before a pull request is submitted.

AI Summary: This issue proposes enhancing the CLI and modernization pipeline to correctly handle COBOL file paths containing Unicode characters, spaces, and special characters. The goal is to improve usability for international users and support common file naming conventions, ensuring all commands and error messages function as expected.

AI Summary: This issue proposes adding a visual progress indicator to the `modernize <file.cbl>` command. The goal is to display which of the five modernization steps is currently running, show a progress bar or spinner for longer steps, and report the duration of each step in real-time. This will improve the user experience by providing clear feedback during the modernization process.