Release v0.6.1

[marcorosa]

Updated

• update models with reference of official sap note from September (add mistral-medium)

Changed

• do not require mandatory tracing

Fixed

• Fix gptfuzz call

Chore

• Add some optimization strategies to Dockerfile to try to respect the memory limit imposed by gha

[github-actions]

The changes encompass various improvements, optimizations, and updates across backend and frontend components, as well as CI/CD workflows. Key updates include version bumps of dependencies, enhanced Dockerfile optimizations, workflow upgrades for better resource management, additions of new model configurations, and adjustments to tracing logic in status.py. Additionally, there are package updates in package.json for Angular and TypeScript dependencies, allowing for more stable and secure frontend development.

Walkthrough

• Chore: Upgraded setup-uv action from v6 to v7 for optimization of CI/CD pipeline with improved cache handling.
• New Feature: Added setup for kubelogin in the GitHub Actions workflows to facilitate Kubernetes login operations.
• Refactor: Enhanced Dockerfile with aggressive cleanup and optimization techniques to reduce image size and improve build speed.
• New Feature: Updated configurations to include new models like gp-4o-mini and mistralai--mistral-medium-instruct.
• Chore: Updated backend Python dependencies for efficiency and security; version bump to 0.6.1.
• Style: Conditional logic added to status.py to ensure a clean trace handling.
• Chore: Upgraded Angular and TypeScript dependencies in package.json to maintain alignment with newer versions, enhancing stability.

_{Model: gpt-4o | Prompt Tokens: 3151 | Completion Tokens: 273}

[github-actions]

Here's a collaborative code review enhanced with AI insights. These suggestions are recommendations to consider rather than definitive requirements, so please apply what feels most valuable for your project. You remain the expert decision-maker—AI simply provides additional perspective to support your development process.

---

Always critique what AI says. Do not let AI replace YOUR I.
_{Model: anthropic--claude-4-sonnet | Prompt Tokens: 5408 | Completion Tokens: 1112}

[github-actions]

Good upgrade to setup-uv@v7 and enabling cache pruning. Consider pinning to a specific version instead of using 'latest' for better reproducibility:

uses: astral-sh/setup-uv@v7
with:
  version: "0.5.4"  # or specific version
  enable-cache: true
  prune-cache: true

[github-actions]

The kubelogin setup step is duplicated across multiple jobs. Consider extracting this into a reusable composite action to reduce duplication:

# .github/actions/setup-kubelogin/action.yml
name: 'Setup kubelogin'
description: 'Sets up kubelogin for Kubernetes authentication'
runs:
  using: 'composite'
  steps:
    - name: Set up kubelogin
      uses: azure/use-kubelogin@v1
      env:
        GITHUB_TOKEN: ${{ github.token }}
      with:
        kubelogin-version: 'latest'

[github-actions]

The Dockerfile improvements look good with aggressive cleanup. However, the cleanup commands could be simplified and made more robust:

# Combine cleanup operations and use proper error handling
RUN . ~/.cargo/env && \
    uv sync --frozen --no-dev --no-cache --compile-bytecode && \
    # Single comprehensive cleanup
    find /app/.venv \( -name "__pycache__" -o -name "*.pyc" -o -name "*.pyo" \) -delete && \
    find /app/.venv -type d \( -name "tests" -o -name "test" -o -name "docs" \) -exec rm -rf {} + && \
    find /app/.venv -name "*.so" -exec strip --strip-debug {} + && \
    rm -rf /root/.cache /tmp/* /var/tmp/* /root/.cargo/registry /root/.cargo/git /app/.venv/share && \
    rustup self uninstall -y && \
    apt-get autoremove -y build-essential git curl pkg-config && \
    apt-get autoclean

[github-actions]

Good defensive programming with the trace check. Consider using a more explicit approach with a proper initialization:

def __init__(self):
    self.trace = {'llm_messages': []}  # Initialize properly
    
def log_llm_message(self, prompt: str, response):
    message = {
        'prompt': prompt,
        'response': response.to_dict()
    }
    # More explicit check
    if self.trace and 'llm_messages' in self.trace:
        self.trace['llm_messages'].append(message)

[github-actions]

Dependency updates look good. Consider using version ranges for better compatibility and easier maintenance:

dependencies = [
    'sap-ai-sdk-gen[all]>=5.7.5,<6.0.0',
    'pandas>=2.3.3,<3.0.0',
    'sentence-transformers>=5.1.1,<6.0.0',
    'tensorflow>=2.20.0,<3.0.0',
    # ... other dependencies
]