Skip to content

Megaphone

v1.1.2 Breaking

This release includes breaking changes for platform teams planning a safe upgrade.

✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

apple apple-ai freeflow macos megaphone megaphone-tts
+5 more
speechanalyzer spokenly text-to-speech whisper whisprflow

Summary

AI summary

Added a personal dictionary with manual entry and on-device term suggestion learning.

Changes in this release

Feature Medium

Adds personal Dictionary for names, products, acronyms, and technical terms with manual entry, search, enable/disable controls, and exact heard-to-written corrections.

Adds personal Dictionary for names, products, acronyms, and technical terms with manual entry, search, enable/disable controls, and exact heard-to-written corrections.

Source: llm_adapter@2026-07-17

Confidence: high

Feature Medium

Adds conservative on-device learning that suggests unusual terms after three observations, with accept/dismiss/disable/remove options.

Adds conservative on-device learning that suggests unusual terms after three observations, with accept/dismiss/disable/remove options.

Source: llm_adapter@2026-07-17

Confidence: high

Feature Medium

Migrates existing Custom Vocabulary entries automatically into the new Dictionary, preserving functionality for SpeechAnalyzer and Smart Cleanup.

Migrates existing Custom Vocabulary entries automatically into the new Dictionary, preserving functionality for SpeechAnalyzer and Smart Cleanup.

Source: llm_adapter@2026-07-17

Confidence: high

Bugfix Medium

Fixes in-app updates to work with macOS App Management, validate publisher/version, retain verified rollback copy, and write durable recovery logs.

Fixes in-app updates to work with macOS App Management, validate publisher/version, retain verified rollback copy, and write durable recovery logs.

Source: llm_adapter@2026-07-17

Confidence: high

Bugfix Medium

Improves Dictionary learning to ignore failed cleanup, commands, macros, Exact mode, common prose, standalone numbers, and dismissed suggestions.

Improves Dictionary learning to ignore failed cleanup, commands, macros, Exact mode, common prose, standalone numbers, and dismissed suggestions.

Source: llm_adapter@2026-07-17

Confidence: high

Full changelog

[1.1.2] - 2026-07-14

Added

  • A personal Dictionary for names, products, acronyms, and technical terms, with manual entry, search, enable/disable controls, and exact heard-to-written corrections.
  • Conservative on-device learning that suggests unusual terms from successful dictations and activates them after three observations. Suggestions can be accepted, dismissed, disabled, or removed at any time.
  • Existing Custom Vocabulary entries migrate automatically into the Dictionary and continue steering both SpeechAnalyzer and Smart Cleanup.

Changed

  • The README demo now uses a real high-resolution recording across Claude Code, Codex, Cursor CLI, and Google Docs, with production cuts and a focused Megaphone overlay animation.

Fixed

  • In-app updates now work with macOS App Management, validate the downloaded publisher and version, retain a verified rollback copy, and write durable recovery logs.
  • Dictionary learning ignores failed cleanup, commands, macros, Exact mode, common prose, standalone numbers, and dismissed suggestions so recognition does not reinforce one-off mistakes.

Download

Megaphone.dmg — requires macOS 26 (Tahoe) on Apple silicon.

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Track Megaphone

Get notified when new releases ship.

Sign up free

About Megaphone

All releases →

Related context

Earlier breaking changes

  • v1.1.0 Edit Mode disabled due to retired LLM layer; setup step removed.
  • v1.1.0 Menu bar icon now hidden by default; must re‑enable in Settings.
  • v1.0.3 Removes API key and provider configuration UI from Settings.
  • v1.0.0 Renamed project from FreeFlow to Megaphone.
  • v1.0.0 Transcription runs entirely on‑device using Apple's SpeechAnalyzer; cloud transcription stack removed.

Beta — feedback welcome: [email protected]