Voice dictation, repo loading, and Code content.

Added a North America Molten Hub OpenAPI schema artifact so agent and integration workflows can inspect the Hub API contract directly.

Improved voice dictation reliability with speech server reachability checks, configurable language and timeout handling, richer audio diagnostics, empty transcript reasons, and lower-signal samples still routed through transcription.

Brought voice dictation into chat repository prompts with microphone capture, waveform status, transcription insertion, and safer messaging for unavailable speech services.

Loaded the full GitHub repository list on app startup and kept repo search, pagination, owner avatars, and the All repositories tab in sync as users switch views.

Hardened no-change follow-ups by archiving local task logs, carrying request, workspace, branch, repo, and target-subdir context into remediation prompts, and skipping escalations when logs already cite concrete no-op evidence.

Published a Molten Hub Code voice dictation setup guide with Docker Compose, Faster Whisper environment settings, UI walkthrough images, and troubleshooting checks.

Published voice dictation and AI-augmented engineering blog posts covering faster repo-change prompts, reviewable transcriptions, autonomy boundaries, and technology-leadership risk mapping.

Added a voice dictation section to the Molten Hub Code page with blog and setup links, updated structured data, responsive styling, and unit coverage.