| name | lightflow-video-work-speech |
| description | This skill should be used when the user asks to "generate speech from a voice profile", "synthesize speech", "use a cloned voice", or "run lightflow.video_work_speech". |
| version | 0.5.0 |
Video Work speech generation
Run lightflow.video_work_speech only with the profile_id returned by a
completed, consent-gated lightflow.video_work_voice_profile run. Treat that
lineage as an operating policy: this workflow validates only a non-empty ID.
Provide non-empty speaker_id, profile_id, and target_text. Keep speed
within the workflow's accepted 0.75 through 1.25 range.
Configure LIGHTFLOW_VIDEO_WORK_API_URL and either
LIGHTFLOW_VIDEO_WORK_API_TOKEN or LIGHTFLOW_VIDEO_WORK_API_TOKEN_FILE only
in the workflow runtime environment. A non-empty token environment value takes
precedence; otherwise the file must be a regular UTF-8 file no larger than 4
KiB with non-empty trimmed contents. Never place or print the URL, token, token
file path, or file contents in workflow inputs, artifacts, source files,
prompts, logs, or command output. Require an HTTPS endpoint unless it targets
loopback.
The service action generate_speech returns a completed generation, not a
queued receipt. The workflow accepts it only when id, audio_url, and
audio_path are non-empty, status is complete, and audio is an object.
Use that validated generation JSON as the audio-production evidence.
CLI Usage
lfw run lightflow.video_work_speech \
--input speaker_id='"narrator-01"' \
--input profile_id='"profile-abc123"' \
--input target_text='"Text to synthesize with the approved voice profile."'
API Usage
curl -sS -X POST http://127.0.0.1:5174/workflows/lightflow.video_work_speech/run \
-H 'content-type: application/json' \
-d '{"inputs":{"speaker_id":"narrator-01","profile_id":"profile-abc123","target_text":"Text to synthesize with the approved voice profile."}}'