Skip to content

Clarify transcribe/stream CLI help text #24

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
alexkroman merged 3 commits into from
Jun 6, 2026
Merged

Clarify transcribe/stream CLI help text #24

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6aaaec4
Clarify transcribe/stream CLI help text
claude Jun 6, 2026
484831d
Merge branch 'main' into claude/cli-help-text-clarity-VjJoN
alexkroman Jun 6, 2026
97e0c3b
Merge branch 'main' into claude/cli-help-text-clarity-VjJoN
alexkroman Jun 6, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 16 additions & 9 deletions aai_cli/commands/stream.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,20 +66,22 @@ def stream(
speech_model: str = typer.Option(
DEFAULT_SPEECH_MODEL, "--speech-model", help="Streaming speech model."
),
encoding: str | None = typer.Option(None, "--encoding", help="pcm_s16le or pcm_mulaw."),
encoding: str | None = typer.Option(
None, "--encoding", help="Audio encoding: pcm_s16le or pcm_mulaw."
),
language_detection: bool | None = typer.Option(
None, "--language-detection", help="Auto-detect the spoken language."
),
domain: str | None = typer.Option(None, "--domain", help="Domain preset (e.g. medical)."),
# turn detection
end_of_turn_confidence_threshold: float | None = typer.Option(
None, "--end-of-turn-confidence-threshold", help="0-1 end-of-turn confidence."
None, "--end-of-turn-confidence-threshold", help="End-of-turn confidence (0-1)."
),
min_turn_silence: int | None = typer.Option(
None, "--min-turn-silence", help="Min turn silence (ms)."
None, "--min-turn-silence", help="Min silence to end a turn (ms)."
),
max_turn_silence: int | None = typer.Option(
None, "--max-turn-silence", help="Max turn silence (ms)."
None, "--max-turn-silence", help="Max silence before ending a turn (ms)."
),
vad_threshold: float | None = typer.Option(
None, "--vad-threshold", help="Voice-activity threshold."
Expand All @@ -99,7 +101,9 @@ def stream(
),
speaker_labels: bool | None = typer.Option(None, "--speaker-labels", help="Label speakers."),
max_speakers: int | None = typer.Option(None, "--max-speakers", help="Max speakers."),
voice_focus: str | None = typer.Option(None, "--voice-focus", help="near_field or far_field."),
voice_focus: str | None = typer.Option(
None, "--voice-focus", help="Voice focus: near_field or far_field."
),
voice_focus_threshold: float | None = typer.Option(
None, "--voice-focus-threshold", help="Voice-focus threshold."
),
Expand All @@ -108,7 +112,7 @@ def stream(
None, "--redact-pii-policy", help="Comma-separated PII policies."
),
redact_pii_sub: str | None = typer.Option(
None, "--redact-pii-sub", help="hash or entity_name."
None, "--redact-pii-sub", help="Replace redacted PII with: hash or entity_name."
),
inactivity_timeout: int | None = typer.Option(
None, "--inactivity-timeout", help="Auto-close after N seconds idle."
Expand All @@ -125,7 +129,9 @@ def stream(
None, "--config-file", help="JSON file of streaming fields."
),
# existing
prompt: str | None = typer.Option(None, "--prompt", help="Bias the speech model (u3-pro)."),
prompt: str | None = typer.Option(
None, "--prompt", help="Prompt to bias the speech model (u3-pro)."
),
llm_prompt: list[str] | None = typer.Option(
None,
"--llm",
Expand All @@ -150,8 +156,9 @@ def stream(
) -> None:
"""Transcribe live audio in real time — from your mic, a file, or a URL.

--prompt biases the speech model. --llm-gateway-prompt transforms the full
transcript through LLM Gateway once the stream ends (e.g. "summarize the call").
--prompt biases the speech model. --llm runs a prompt over the live transcript
through LLM Gateway, refreshing the answer on every finalized turn (e.g.
"summarize action items").
"""

def body(state: AppState, json_mode: bool) -> None:
Expand Down
24 changes: 14 additions & 10 deletions aai_cli/commands/transcribe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -82,7 +82,7 @@ def transcribe(
sample: bool = typer.Option(False, "--sample", help="Use the hosted wildfires.mp3 sample."),
# model & language
speech_model: str | None = typer.Option(
None, "--speech-model", help="best, nano, slam-1, universal."
None, "--speech-model", help="Speech model: best, nano, slam-1, or universal."
),
language_code: str | None = typer.Option(
None, "--language-code", help="Force a language (e.g. en_us)."
Expand All @@ -96,22 +96,26 @@ def transcribe(
temperature: float | None = typer.Option(
None, "--temperature", help="Speech model temperature."
),
prompt: str | None = typer.Option(None, "--prompt", help="Bias the speech model (u3-pro)."),
prompt: str | None = typer.Option(
None, "--prompt", help="Prompt to bias the speech model (u3-pro)."
),
# formatting
punctuate: bool | None = typer.Option(
None, "--punctuate/--no-punctuate", help="Add punctuation."
),
format_text: bool | None = typer.Option(
None, "--format-text/--no-format-text", help="Format text."
None, "--format-text/--no-format-text", help="Apply text formatting (casing, numbers)."
),
disfluencies: bool | None = typer.Option(
None, "--disfluencies", help="Keep filler words (e.g. um, uh)."
),
disfluencies: bool | None = typer.Option(None, "--disfluencies", help="Keep filler words."),
# speakers & channels
speaker_labels: bool = typer.Option(False, "--speaker-labels", help="Enable diarization."),
speakers_expected: int | None = typer.Option(
None, "--speakers-expected", help="Hint speaker count."
),
multichannel: bool | None = typer.Option(
None, "--multichannel", help="Transcribe each channel."
None, "--multichannel", help="Transcribe each audio channel separately."
),
# guardrails
redact_pii: bool | None = typer.Option(
Expand All @@ -121,7 +125,7 @@ def transcribe(
None, "--redact-pii-policy", help="Comma-separated PII policies (e.g. person_name,...)."
),
redact_pii_sub: str | None = typer.Option(
None, "--redact-pii-sub", help="Substitution: hash or entity_name."
None, "--redact-pii-sub", help="Replace redacted PII with: hash or entity_name."
),
redact_pii_audio: bool | None = typer.Option(
None, "--redact-pii-audio", help="Also redact audio."
Expand All @@ -133,20 +137,20 @@ def transcribe(
None, "--content-safety", help="Detect sensitive content."
),
content_safety_confidence: int | None = typer.Option(
None, "--content-safety-confidence", help="Confidence threshold 25-100."
None, "--content-safety-confidence", help="Content-safety confidence threshold (25-100)."
),
speech_threshold: float | None = typer.Option(
None, "--speech-threshold", help="Minimum speech proportion 0-1."
None, "--speech-threshold", help="Minimum proportion of speech required (0-1)."
),
# analysis
summarization: bool | None = typer.Option(
None, "--summarization", help="Summarize the transcript."
),
summary_model: str | None = typer.Option(
None, "--summary-model", help="informative/conversational/catchy."
None, "--summary-model", help="Summary model: informative, conversational, or catchy."
),
summary_type: str | None = typer.Option(
None, "--summary-type", help="bullets/gist/headline/paragraph."
None, "--summary-type", help="Summary format: bullets, gist, headline, or paragraph."
),
auto_chapters: bool | None = typer.Option(None, "--auto-chapters", help="Generate chapters."),
sentiment_analysis: bool | None = typer.Option(
Expand Down
76 changes: 48 additions & 28 deletions tests/__snapshots__/test_cli_output_snapshots.ambr
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -517,9 +517,9 @@

Transcribe live audio in real time — from your mic, a file, or a URL.

--prompt biases the speech model. --llm-gateway-prompt transforms the full
transcript through LLM Gateway once the stream ends (e.g. "summarize the
call").
--prompt biases the speech model. --llm runs a prompt over the live transcript
through LLM Gateway, refreshing the answer on every finalized turn (e.g.
"summarize action items").

╭─ Arguments ──────────────────────────────────────────────────────────────────╮
│ source [SOURCE] Audio file path, URL, or YouTube URL to stream. Omit │
Expand All @@ -546,18 +546,19 @@
│ --speech-model TEXT Streaming speech │
│ model. │
│ [default: u3-rt-pro] │
│ --encoding TEXT pcm_s16le or │
│ --encoding TEXT Audio encoding: │
│ pcm_s16le or │
│ pcm_mulaw. │
│ --language-detection Auto-detect the │
│ spoken language. │
│ --domain TEXT Domain preset (e.g. │
│ medical). │
│ --end-of-turn-confi… FLOAT 0-1 end-of-turn │
│ confidence. │
│ --min-turn-silence INTEGER Min turn silence │
│ (ms). │
│ --max-turn-silence INTEGER Max turn silence │
│ (ms). │
│ --end-of-turn-confi… FLOAT End-of-turn │
│ confidence (0-1). │
│ --min-turn-silence INTEGER Min silence to end a │
│ turn (ms). │
│ --max-turn-silence INTEGER Max silence before │
│ ending a turn (ms). │
│ --vad-threshold FLOAT Voice-activity │
│ threshold. │
│ --format-turns --no-format-turns Punctuate/format │
Expand All @@ -568,15 +569,18 @@
│ --filter-profanity Mask profanity. │
│ --speaker-labels Label speakers. │
│ --max-speakers INTEGER Max speakers. │
│ --voice-focus TEXT near_field or │
│ --voice-focus TEXT Voice focus: │
│ near_field or │
│ far_field. │
│ --voice-focus-thres… FLOAT Voice-focus │
│ threshold. │
│ --redact-pii Redact PII from │
│ turns. │
│ --redact-pii-policy TEXT Comma-separated PII │
│ policies. │
│ --redact-pii-sub TEXT hash or entity_name. │
│ --redact-pii-sub TEXT Replace redacted PII │
│ with: hash or │
│ entity_name. │
│ --inactivity-timeout INTEGER Auto-close after N │
│ seconds idle. │
│ --webhook-url TEXT Webhook URL. │
Expand All @@ -588,8 +592,9 @@
│ (repeatable). │
│ --config-file TEXT JSON file of │
│ streaming fields. │
│ --prompt TEXT Bias the speech │
│ model (u3-pro). │
│ --prompt TEXT Prompt to bias the │
│ speech model │
│ (u3-pro). │
│ --llm TEXT Run a prompt over │
│ the live transcript │
│ through LLM Gateway, │
Expand Down Expand Up @@ -658,7 +663,8 @@
│ --sample Use the hosted │
│ wildfires.mp3 │
│ sample. │
│ --speech-model TEXT best, nano, slam-1, │
│ --speech-model TEXT Speech model: best, │
│ nano, slam-1, or │
│ universal. │
│ --language-code TEXT Force a language │
│ (e.g. en_us). │
Expand All @@ -668,34 +674,48 @@
│ (repeatable). │
│ --temperature FLOAT Speech model │
│ temperature. │
│ --prompt TEXT Bias the speech │
│ model (u3-pro). │
│ --prompt TEXT Prompt to bias the │
│ speech model │
│ (u3-pro). │
│ --punctuate --no-punctuate Add punctuation. │
│ --format-text --no-format-text Format text. │
│ --disfluencies Keep filler words. │
│ --format-text --no-format-text Apply text │
│ formatting (casing, │
│ numbers). │
│ --disfluencies Keep filler words │
│ (e.g. um, uh). │
│ --speaker-labels Enable diarization. │
│ --speakers-expected INTEGER Hint speaker count. │
│ --multichannel Transcribe each │
│ channel. │
│ audio channel │
│ separately. │
│ --redact-pii Redact PII from the │
│ transcript. │
│ --redact-pii-policy TEXT Comma-separated PII │
│ policies (e.g. │
│ person_name,...). │
│ --redact-pii-sub TEXT Substitution: hash │
│ or entity_name. │
│ --redact-pii-sub TEXT Replace redacted PII │
│ with: hash or │
│ entity_name. │
│ --redact-pii-audio Also redact audio. │
│ --filter-profanity Mask profanity. │
│ --content-safety Detect sensitive │
│ content. │
│ --content-safety-con… INTEGER Confidence threshold │
│ 25-100. │
│ --speech-threshold FLOAT Minimum speech │
│ proportion 0-1. │
│ --content-safety-con… INTEGER Content-safety │
│ confidence threshold │
│ (25-100). │
│ --speech-threshold FLOAT Minimum proportion │
│ of speech required │
│ (0-1). │
│ --summarization Summarize the │
│ transcript. │
│ --summary-model TEXT informative/convers… │
│ --summary-type TEXT bullets/gist/headli… │
│ --summary-model TEXT Summary model: │
│ informative, │
│ conversational, or │
│ catchy. │
│ --summary-type TEXT Summary format: │
│ bullets, gist, │
│ headline, or │
│ paragraph. │
│ --auto-chapters Generate chapters. │
│ --sentiment-analysis Analyze sentiment. │
│ --entity-detection Detect entities. │
Expand Down
Loading