Changes
1.7.0 (2026-03-05)
- Feature: Streaming Endpoint Refactor - Switched to POST-based fetch with ReadableStream for better reliability and to avoid URL length limits.
- Feature: Real-time Streaming Status - Added status indicators (Connecting, Receiving, Done) in the AI response modal.
- Feature: Added support for Gemini, Claude, DeepSeek, and Custom OpenAI-compatible endpoints.
- Fix: Permission issues resolved for all plugin files.
- Fix: ESLint and AMD build compliance.
1.6.0 (2026-03-04)
- Feature: Added support for Gemini API (Google).
- Feature: Added support for Claude API (Anthropic).
1.5.0 (2026-01-31)
- Feature: Dynamic Prompt Generation - prompt updates in real-time as you type, removing the need for a "Generate" button.
- Refactor: Removed "Insert into Editor" functionality to simplify the block and decouple it from specific editors.
- Fix: Improved Ollama streaming stability with timeout handling and immediate availability of response actions.
- Fix: Removed manual "Generate prompt" button for a smoother workflow.
1.4 (2026-01-31)
- Fix: Resolved PHP TypeError (htmlspecialchars) when rendering course competencies in Mustache templates.
- Fix: Achieved full ESLint compliance for all Javascript modules (actions, ui, markdown, stream, etc.).
- Refactor: Split complex logic in
markdown.js and ui.js into smaller, maintainable helper functions.
- Clean: Removed trailing whitespace and addressed all pending linting warnings.
1.3 (2026-01-31)
- Feature: Refined AI Response Modal with 4 view modes: RAW, TEXT, HTML (Rich preview), and HTML CODE (source).
- Feature: Smart "Copy to Clipboard" with support for copying rich text.
- Feature: Improved UI feedback with copy status indicators.
- Feature: Enhanced Markdown rendering compatibility.
- Fix: Improved auto-formatting for streamed AI responses.
- Fix: UI layout refinements for better modal visibility.
1.2 (2025-12-06)
- Feature: Added RAW, TEXT, and HTML view modes for AI response improved accessibility and usage.
- Feature: Added "Copy" buttons that respects current view mode.
- Feature: Client-side Markdown rendering with auto-fix for clumped text from streams.
- Fix: Ollama connection now supports local/private IP addresses by bypassing SSL verification for local subnets.
- Fix: Reduced modal height to prevent buttons from being pushed off-screen.
- Fix: Improved TEXT view cleanup (removed stray markdown artifacts).
1.0 (2025-12-04)
- Stable release.
- Added unified "Send to AI" workflow.
- Added Ollama provider support.
- Refined prompt builder UI.
0.3.0 (2025-08-31)
- Unified "Send to AI" button with provider select.
- Added Ollama endpoint & model settings.
0.2.0 (2025-08-25)
- Initial OpenAI ChatGPT send support.
- Prompt builder refinements.
0.1.0
- First public prototype (prompt builder only).