Skip to content

Call-record sidecars for tts, speech_clone, speech_design#6

Merged
TroyHernandez merged 4 commits into
mainfrom
call-sidecars
Jun 12, 2026
Merged

Call-record sidecars for tts, speech_clone, speech_design#6
TroyHernandez merged 4 commits into
mainfrom
call-sidecars

Conversation

@TroyHernandez

Copy link
Copy Markdown
Contributor

tts.api leg of the provenance arc (same convention as cornball-ai/xtx.api#11): every audio asset gets <file>.json recording the resolved call (voice, model, backend, ...) under the cornball_sidecar v1 schema. One .sidecar_arm(environment(), "file") line per public function; the on.exit hook writes only when the output file exists with a fresh mtime, so error paths and skipped work record nothing. 54/54 tests.

Test input validation for voice cloning and voice design functions.
Verify tts_providers structure and expected provider entries.
Brings test count from 11 to 41.
Every audio asset gets <file>.json recording the resolved call (voice,
model, backend, ...) per the cornball_sidecar v1 convention shared with
xtx.api: an on.exit hook writes the record only when the output exists
with a fresh mtime. Sidecars ride with the media and survive bundle
rebuilds; cornductor scoops them into OTIO media-reference metadata.
@TroyHernandez TroyHernandez merged commit a3f7842 into main Jun 12, 2026
4 checks passed
@TroyHernandez TroyHernandez deleted the call-sidecars branch June 12, 2026 00:26
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant