Closes #47 #48

Fixed context caching and Improved LLM prompt handling, and response processing

Changes

• Updated inline templates to explicitly indicate cursor position and prevent modification of existing code.
• Enhanced cleanSuggestion to retain the original suggestion when the cleaned suggestion is empty.
• Fixed bug in API endpoint handling in Gemini, Mistral, and OpenAI providers.
• Increased timeout for robustFetch to handle longer network delays.
• Updated cache key generation to include context documents.
• Improved logging clarity and corrected cache key generation in llmManager.

Flags

• Needs testing.

Screenshots or Video

Related Issues

• Issue Context File Caching Does Not Reflect Recent Changes #47 OpenAI and Mistral Embeddings Functionality Verification Needed #48
• Pull Request #

Author Checklist

• Ensure you provide a DCO sign-off for your commits using the --signoff option of git commit.
• Vital features and changes captured in unit and/or integration tests
• Commits messages follow AP format
• Extend the documentation, if necessary
• Merging to master from fork:branchname