Voice Append 1.0.2

Voice Append 1.0.2 makes provider key selection more reliable and integrates Obsidian’s built-in keychain picker.

• Select an existing Obsidian secret or create a new one for the LLM or transcription provider.
• Keep each device’s secret selection local, so synced plugin settings cannot replace it.
• Configuration tests no longer save empty or stale API key inputs.
• Legacy key recovery no longer mistakes a separate transcription key for the main provider key.
• Existing Voice Append keys remain usable automatically. A missing or cleared selection never silently falls back to another credential.

After updating, open Settings → Voice Append to select or create a provider secret if needed. Configure the key separately on each device; secret values and selections are not synced with the vault.

Requires Obsidian 1.11.4 or later. Validation: 74 automated tests, lint, TypeScript checks, and production build passed.

Voice Append 1.0.1

This maintenance release resolves the Community Plugins review findings.

• Makes production builds deterministic so release assets match a clean repository build.
• Adds searchable settings definitions for Obsidian 1.13 and later.
• Removes the one-time full-vault migration scan.
• Uses window-scoped timers for pop-out window compatibility.

All 59 automated tests pass.

Voice Append 1.0.0

Voice Append 1.0.0 is the first public release.

Highlights

• Record from the end of a note, the command palette, the ribbon, or Obsidian's mobile toolbar.
• Save recordings locally before processing and retry them after restarts or network failures.
• Configure shared, separate, or local OpenAI-compatible transcription and LLM providers.
• Test the complete provider configuration with a bundled sample recording.
• Clean up transcripts with configurable instructions, optional note context, and familiar terminology.
• Generate filenames for empty notes while preserving Unique Notes timestamps when desired.
• Keep the screen awake while recording where the platform permits it.
• Use localized English and German controls on mobile and desktop.

API keys remain in Obsidian Secret Storage. Voice Append contains no telemetry and sends data only to the providers configured by the user.

See the README for setup, privacy details, limitations, and manual installation instructions.