Kovati Docs

SpeechForgeToolset

Every released version of SpeechForgeToolset, newest first. Latest 0.2.2.

SpeechForgeToolset0.2.2Open source
SetSpeechForge
Roletoolset
Latest0.2.2 — 11 September 2026
Versions here4
Depends onSpeechForge
Sourcegithub.com/AutomationForgeHQ/SpeechForgeToolset

0.2.x

0.2.2

11 September 2026 · Download (181 KB) · Release on GitHub

Added

  • ProduceSpeechBank — produce a whole bank the bank's own way, the agent-side twin of the Speech Library's Generate All. On a plain bank that generates what is missing or stale; on a localised bank it generates the synthesis-sourced lines, dubs the performed ones, and points the face banks serving it at the new audio.

Changed

  • The version this toolset reports to an agent is now read from the plugin's own descriptor rather than repeated in C++, so it can no longer answer a number the installed package does not carry.

0.2.1

8 September 2026 · Download (174 KB) · Release on GitHub

  • Packaging fix: the release now carries everything the register allows. BuildPlugin's filter excludes Config/ and every public_extra path, so earlier zips shipped without them.
  • GetToolsetVersion() answers this plugin's real version; it had drifted from the descriptor.

0.2.0

7 September 2026 · Download (171 KB) · Release on GitHub

  • Apache-2.0 relicensing; a release now publishes its source
  • Every plugin descriptor made to agree with its release tag and credit its author
  • The culmination wiring: one library, a teleprompter, and panels that meet
  • Speakers, voice profiles, and the three-page Speech Library
  • Sessions and takes: recording stops deciding, the director chooses
  • Documentation pass to catch up with the code
  • ConvertSpeechLine shipped — re-voice a line without rewriting it
  • Cost estimation extended to the conversion pipeline
  • Subtitle-worthiness asked of every line, not just tool-fixable ones
  • A re-voicing given its own path, distinct from a new take
  • Fixed: pressing Use gave a different voice than expected, and billed for it
  • Takes always remembered; "in use" now measured rather than claimed
  • Save Takes made a latch; takes can be removed or brought back in
  • Speech and face joined by id: a speaker names their head, a bank says what it holds
  • Localisation: a sibling bank per language, joined by line id
  • Repointed to kovati.dev

0.1.x

0.1.0

28 August 2026 · Download (118 KB) · Release on GitHub

  • Initial release: the pipeline as MCP tools
  • Output roots reworked so generated assets move independently of their sources

On this page