Feature Catalog

Feature Catalog

YazSes ships 139 capabilities across 9 areas — from core dictation to accessibility, meetings, and on-device learning. Almost everything is opt-in and off by default; a fresh install is simple, and you turn on only what you want with yazses features enable <name>.

This page is a scannable map for contributors. For the exhaustive per-feature reference (toggles, config keys, examples) see the feature reference on the docs site.

Want to help build one of these? Every area below has a "good ways to contribute" note. Pick an area in the 🧩 Adopt a feature discussion, grab a good first issue, or open an 💡 Idea.

Areas at a glance

Area	Features
Core dictation	19
Accuracy & correction	20
Formatting & structure	31
Editing & navigation	13
Commands & automation	12
Multilingual	7
Accessibility & input modalities	20
Learning, memory & analytics	9
Conversation & recording capture	8

Core dictation

The heart of YazSes. Great for contributors who care about latency, capture reliability, and the hold-to-talk UX. Low-level audio/VAD and hotkey work lives here.

Feature	What it does
Dictation core	The core hold-to-talk transcription.
Voice commands	Spoken commands like 'undo', 'save', 'delete last word'.
Voice punctuation	Say 'comma', 'period', 'new line', 'question mark' to insert marks.
Mid-Thought Undo	Say 'scratch that' to drop the last phrase.
Voice-activity overlay	Sonar rings near the cursor while you talk.
System-tray icon	A microphone icon in the top bar; click it to pick/pin your mic, re-calibrate, and start/stop the daemon.
Text-target guard	When you dictate with no text field focused, don't type into the wrong place — turn the tray icon yellow and copy the transcript to the clipboard instead.
Mic-change guard	Notifies + auto-heals when your microphone silently switches (e.g. a USB-C monitor stealing capture) so dictation never dies in silence.
Ghost Ahead	Pre-warms the decoder for slightly faster first words.
Streaming transcription	Injects words live as you speak (overtype).
Accessibility Continuum	Whisper/Low-Effort Mode lowers the mic gate so quiet or effortful speech is still captured (no shouting).
Wake-Word Activation	Start dictation hands-free by saying a keyword.
Breath-Paced Dictation	Uses your natural breath onsets as sentence/paragraph boundaries, so dictation chunks the way you actually breathe — even without silent pauses.
Whisper-Aware Mode	Detects when you drop to a whisper and adapts gain, VAD, and the STT prompt so quiet dictation stays accurate in shared or private spaces.
Hesitation-Hold Endpointing	Holds the turn open when you trail off with a filled hesitation ('annnd… uhh…') instead of cutting you off — for anyone who thinks out loud.
Focus-Class Auto-Profile	Auto-selects the dictation profile from the focused window's class (terminal→shell, editor→code, browser→prose) so you never switch profiles by hand.
Adaptive Latency Governor	Keeps dictation responsive under CPU load (lighter model/beam) and losslessly faster when idle (speculative decoding).
Hands-Free Auto-Stop	Tap once and speak; recording auto-stops when you finish (silence timeout + duration cap).
Codec Streaming (low latency)	Routes decoding to a streaming neural-codec engine for lower latency.

Accuracy & correction

Where transcription becomes right. Ideal for people into ASR, error-correction heuristics, and evaluation. Many pieces are pure Python you can improve with a test.

Feature	What it does
Dysfluency-Friendly	Collapses stutters/repeats (b-b-because→because).
Punch-In	Re-speak a phrase to correct the last one.
Confidence Ink	Marks words Whisper was unsure about so you can re-pick them by voice instead of re-dictating.
Context-Primed Dictation	Primes STT with terms from the active window/selection so domain words are transcribed right.
Diacritize	Restore the diacritics ASR drops — 'cafe' → 'café', 'naive' → 'naïve' — via an unambiguous lexicon (ambiguous words left alone).
SafeGlyph	Flag Unicode confusables in dictated identifiers/URLs — Cyrillic 'а' vs Latin 'a', zero-width characters, mixed-script words — before they are injected.
Checksum-Validated Entry	Runs dictated account/ID numbers (credit card, IBAN, ISBN) through their check-digit algorithm and flags a mis-heard digit before it lands — with fix suggestions.
Involuntary-Vocalization Excision	Detects and deletes involuntary sounds (cough, throat-clear, sneeze) from the stream so they never become garbage tokens or spurious commands.
Self-Learning Correction Dictionary	Learns the ASR errors you keep fixing (e.g. 'yaz says' → 'YazSes') and applies them automatically to future output.
Confidence-Gated Re-Ask	Instead of injecting a low-confidence guess, holds just the uncertain word and asks you to pick or repeat it — closing the correction loop.
Phonetic Spelling Mode	A character-exact mode for passwords/codes/IDs: 'capital alpha bravo double lima' → 'Abll'.
Hard Contextual Biasing	Biases recognition toward your vocabulary with a hotword trie (not just a soft prompt), so rare names and jargon are transcribed right.
Mid-Utterance Self-Repair	Applies in-burst corrections before typing: 'email Sarah no I mean Sara' → 'email Sara'.
Screen-Grounded Dictation	Primes Whisper with names visible on screen so they transcribe right the first time — works in any app.
Grammar Repair	Fixes article/agreement errors in dictated text with minimal edits, tuned for non-native speakers.
Acoustic Profiles	Detects your environment (quiet/café/car/meeting) and auto-tunes the mic gate + noise suppression.
Audio-Anchored Scrubbing	Keeps word-level timestamps so you can replay what you said or pick a word to re-dictate just that word.
Phonetic Corrector	Fixes mis-heard proper nouns by sound against your vocabulary ('Cuber Netties'->'Kubernetes').
Hallucination Guard	Drops Whisper's fabricated text on silence/noise (the phantom 'Thank you.', 'please subscribe', or looped phrases) before it's typed.
Noise Suppression	Removes background noise/echo before transcription so dictation works in noisy rooms.

Formatting & structure

Turning a stream of words into well-formed text/code/tables. A huge surface of small, self-contained, testable features — the richest area for first PRs.

Feature	What it does
Prosody Ink	Turns pauses into paragraphs and stressed words into bold.
LLM cleanup	Reformats dictation with a small offline LLM.
BrailleOut	Emit Grade-2 (UEB) Unicode Braille cells so a Braille-display or DeafBlind user gets dictation directly in Braille — no screen-reader round-trip.
Spoken Outline	Say 'new item / indent / promote / collapse' to drive a live outline tree and render it as Markdown or OPML — structured idea capture by voice.
Diagrams-as-Code by Voice	Dictate a flowchart ('start goes to login; login goes to dashboard if success') and get Mermaid/Graphviz source — draw diagrams without a mouse or canvas.
Screenplay Auto-Format	Formats dictated scenes and dialogue as Fountain screenplay markup (scene headings, character cues, smart curly quotes).
Style-Consistency Enforcer	Applies your style sheet to each dictation ('e-mail' not 'email', US spelling, 'cannot' not 'can not') so terminology stays consistent — a Vale-lite pass.
Suggestion-Mode Dictation	Emits dictated edits as CriticMarkup tracked changes ({++add++}, {--cut--}, {~~old~>new~~}) for an editor to review later, instead of applying them.
Semantic Line Breaks	Breaks dictated prose one clause per line so git diffs and merges stay clean — the rendered output is unchanged.
Acronym & Glossary Manager	Expands an acronym on first mention ('World Health Organization (WHO)'), contracts it after, and warns about acronyms you never defined.
Prosodic Auto-Punctuation	Inserts periods, commas, and question marks from how you speak (pauses, pitch) — no need to say 'comma' or 'period'.
Spoken Table Entry	Bulk data entry by voice: 'row: Ada, 1815, London' types tab/comma cells and 'next row' moves down.
Voice Case Transform	Recase the selected text by voice: 'make this snake_case', 'Title Case', 'SHOUT', 'camelCase'.
Auto-Pairing & Wrap	Balances brackets/quotes in dictated code and wraps a selection ('wrap this in parens').
Inline Compute	Say 'what's 15% of 240' and it types 36.
Verbatim / Autoformat Toggle	Say 'dictate verbatim' to freeze all formatting (ITN, punctuation, reflow) for exact capture, 'resume formatting' to restore — mid-burst.
Structured-Markup Dictation	Speak structure and get Markdown/org: 'bullet list: apples; oranges' → a list; 'table columns Name, Age; row Alice, 30' → a table.
Spoken Regex Builder	Build regexes by voice: 'four digits dash two digits' → \d{4}-\d{2}.
On-Device Condense	A condense variant inserts a tightened summary of your long dictation instead of the verbatim text.
Spoken Spreadsheet	Cell-addressed dictation + grid navigation for spreadsheets: 'next row', 'cell down', 'go to B7'.
Spoken Temporal Normalizer	Resolves spoken dates against the clock: 'next Friday' → a concrete date, 'tomorrow', 'in two weeks'.
Emoji & Symbol by Voice	Speak emoji, symbols and arrows: 'shrug emoji' → 🤷, 'right arrow' → →, 'degree sign' → °.
Voice Unit Conversion	Evaluate spoken conversions inline: 'twenty miles in kilometers' → '32.19 kilometers'.
Redaction Ink	Detects spoken secrets (card numbers, SSNs, keys, emails) and masks them before they're typed into another window.
Field-Aware Dictation	Reshapes output to the focused control: number field → digits, search box → no trailing period, password field → refuses to type.
Entity ITN	Turns spoken structured entities into written form with no command words: 'john dot doe at gmail dot com' → the email, 'version two point one' → v2.1.
Dictation Reflow	Say 'structure this' to rewrite your last ramble into bullets and action items.
Smart-Paste	Adapts injected syntax to the target app (markdown bullets, code casing, URL autolinking) using local window info.
Spoken Code Mode	Dictate code: spoken symbols become punctuation and word-groups become cased identifiers (snake/camel/pascal).
Spoken Math (LaTeX)	Dictate math and inject LaTeX ('x squared plus y squared' -> x^{2} + y^{2}).
Tone-Aware Formatting	Adds !

Editing & navigation

Editing by voice ('change X to Y', structural selection). Fun for people who like text-editor internals and command grammars.

Feature	What it does
Spoken Edit Mode	Edit the last dictation by voice ('change their to there', 'delete the last sentence').
HatSelect Structural Editing	Every visible token gets a spoken label so you can edit structurally — 'select alpha to charlie', 'delete bravo' — without cursor motions (Cursorless-style).
WordFind Reverse Dictionary	Offline reverse dictionary for tip-of-the-tongue moments — 'the word for when water turns to gas' → a ranked shortlist (evaporation, …).
Spaced-Repetition Capture	Say 'remember that X is Y' mid-dictation and it makes a local Anki cloze study card, scheduled by SM-2 — study material captured in the flow of work.
Word-Count & Goal Tracker	Counts dictated words and tracks a goal: say 'goal 500 words', then 'how many words so far?' for spoken progress.
Voice Undo/Redo Timeline	Undo/redo YazSes's own output across bursts by voice — 'undo the last sentence', 'redo' — even where the app's Ctrl+Z is unreliable.
Session Bookmarks	Drop named anchors in a long dictation ('bookmark here as intro') and jump back ('jump to my last bookmark') after a break.
Voice Jump-to-Symbol	'Jump to function tokenize', 'go to line 240' → the editor moves there (fuzzy symbol match, or a search fallback).
Voice Fuzzy File Open	'Open the notes about the mortgage' → fuzzy-matches your local files and opens the best one.
Citation-by-Voice	Say 'cite Vaswani 2017' and it inserts a formatted citation from your local BibTeX library, fully offline.
Voice Window Management	Hands-free desktop layout: 'move window left half', 'maximize', 'workspace 3'.
Document Find-and-Replace	Edit the whole document by voice: 'replace every utilise with use'.
Clipboard-History by Voice	Recall recent copies by voice: 'paste the second thing I copied', 'the URL I copied'.

Commands & automation

Voice → actions/macros/shortcuts. Good for automation tinkerers and per-app profile authors.

Feature	What it does
Say-Macro	Speak a trigger word to expand canned text.
Chorded Shortcut Synthesis	Say any shortcut and it's pressed — 'press control shift P', 'escape twice', 'hit F5' — no macro registration needed.
Spoken Shell Pipeline Builder	Speak stages ('list files, pipe to grep error, pipe to word count') → renders 'ls | grep error | wc -l' as text; nothing runs until you say 'run it'.
Voice Git Choreographer	Drive git by voice via a structured grammar; destructive ops (force-push, reset --hard) wait for a spoken confirm and the undo is always spoken.
Terminal Command Safety Gate	In a terminal, holds a destructive command (rm -rf, curl|sh, force-push) until you say 'confirm', so a misrecognition can't fire it.
Slot-Filling Dictation	Speak once and fill a named-field template: 'bug in login, high priority, affects Firefox' routes each value to its field.
Few-Shot Command Spotter	Enroll a few examples of short commands ('send', 'stop') that fire instantly without a full decode.
Ambient Audio-Event Guard	Auto-pauses dictation (or alerts) when it hears a doorbell, phone, alarm or your name.
Gesture Chords	Bind chords (held key + nod / second key / sEMG squeeze) to actions like send or switch profile.
Voice Snippets	Say a trigger ('insert my signature') to type a stored template.
Voice-to-Tool (Spoken MCP)	Speak an intent to run allowlisted tools via MCP; state-changing tools ask first.
Voice Pilot (AT-SPI)	Drive the desktop by voice via the accessibility tree ('click Save', 'focus the terminal').

Multilingual

Non-English & code-switching. High-impact and high-visibility — every language you add reaches a new community. Often docs + config, no deep code.

Feature	What it does
Polyglot Switch	Handles dictation that mixes two languages.
Transliteration	Dictate your native language in Latin letters ('salam, chetori?') and inject the native script — Whisper nails romanized phonetics where native audio fails.
Per-Language Auto Switching	Detects the language you speak and hot-swaps to its specialized model + punctuation, no manual toggle.
Compose-in-Target-Language	Speak your strongest language and inject the text in a target language you write less fluently.
Two-Way Interpreter	Face-to-face mode: two speakers alternate and each turn is translated into the other language.
Pronunciation Feedback	Practice mode: dictate a target phrase and get per-phoneme good/fair/poor feedback for accent training.
Speech Translation	Speak another language and type English (Whisper's built-in translate; no extra download).

Accessibility & input modalities

The features that make YazSes usable for people who can't rely on a keyboard. The most meaningful area to contribute to, and where real-user feedback matters most.

Feature	What it does
Read-Back Loop	Speaks the transcript back to you (accessibility).
LoadGuard	When your speech shows rising cognitive load (pauses, fillers, slowdown), widen confirmations and defer risky actions instead of firing them.
Echo Own-Audio Replay	Say 'play that back' to replay your own captured audio for a text span (not TTS) — catch homophone/ASR errors by ear, eyes-free.
Screen-Reader Pacing	Pace injected text to a screen reader's reading rate, clause-chunked, instead of one burst — so the screen reader can announce it coherently.
Interruptible Proofreading	Interrupt the spoken read-back ('stop — change that') and the cursor lands on the exact word being read — eyes-free proofreading.
Vocal Morse	Type by Morse using two vocal sounds (short/long) — full text from a single reliable vocalization, with adaptive timing.
Mouth-Sound Switch Access	Drive a scanning selector with non-verbal mouth sounds ('pop' = advance, cluck = select) — the switch-access method for people who can't produce reliable speech.
Pitch-Contour Gestures	Hum a pitch shape as a word-free command: rising = confirm, falling = cancel, rise-fall = undo.
Beam-Steered Spatial VAD	With a stereo/2-mic input, drops any sound not coming from your seat (a TV, a colleague) by its arrival direction — enrollment-free, complements Cocktail Filter.
Vocal Joystick	Continuous cursor/scroll control by sustaining vowels — 'ahh' right, 'eee' up, louder = faster, pitch-jump = click.
Earcon Feedback	Tiny structured tones signal daemon state eyes-free (rising motif = recording, buzz = low confidence, chime = command done) — faster than spoken read-back.
Local Voice Timer	Set a timer or break reminder by voice ('set a timer for 25 minutes'), announced by read-back — fully offline, no phone or cloud assistant.
Silent Lip-Reading	Dictate silently — a webcam reads your lips when you can't or shouldn't speak.
Sign-Language Input	Deaf/HoH signers dictate by signing to the webcam; ASL is recognized on-device and typed.
Head-Pointer	Move and click the mouse by tilting your head; pair with voice for a fully hands-free desktop.
Personal Read-Back Voice	Read the transcript back in a clone of your own voice from a short enrollment.
Vocal-Strain Guard	Advises a break when your voice shows rising strain (jitter/shimmer/HNR) over a session.
Voice Mouse Grid	Drive the cursor and click by voice via a numbered grid ('three, seven, click') where no accessibility tree exists.
Modality Role Router	Routes each input to its fastest role (gaze→point, EMG→command, voice→dictation).
Glance-Type (camera)	Uses the webcam to route dictation to the pane you look at.

Learning, memory & analytics

The opt-in, on-device self-improvement loop (encrypted corpus, tuning). For people into privacy-preserving ML and on-device personalization.

Feature	What it does
Voiceprint Mind (personalize)	Biases STT toward terms you use often.
Learning loop	Records an encrypted local corpus so yazses tune can improve accuracy.
Corpus Voiceprint Scrub	Speaker-anonymizes learning-corpus audio before storage so retained clips keep what was said but not an identifiable voice.
Mood Ledger	Tags each dictation with an emotion and builds a private mood-over-time view.
Speaking Coach	Private on-device analytics of your dictation: filler rate, words-per-minute, vocabulary diversity, trend.
Multi-User Profiles	On a shared machine, loads each enrolled speaker's own vocab/hotkey/cleanup from their voiceprint — no manual switching.
Ask My Notes (voice RAG)	Ask a question by voice and get an answer grounded in — and citing — your own local notes/docs.
Predictive Completion	A tiny local model suggests the rest of your sentence; accept by voice.
Spoken Recall & Scratch	Search your past dictations ('yazses recall …') and capture spoken notes-to-self.

Conversation & recording capture

Meetings and offline file transcription with speaker attribution. Good for audio-pipeline and diarization work.

Feature	What it does
Recording Import	Run yazses transcribe <file> to batch-transcribe voice memos/lectures/meetings offline to .txt/.md/.srt/.vtt/.json.
Meeting Mode	yazses meeting start records a whole meeting hands-free (no key to hold), streams a live transcript, and at yazses meeting stop writes a speaker-labelled transcript ('Alice: …', 'You: …') plus opt-in notes.
Crowd-Proof Dictation	Reconstructs your enrolled voice out of overlapping babble before STT, so dictation survives an open-plan office or café.
Diarized Conversation Capture	Capture a live conversation as attributed Markdown (Alice: …) and rename speakers by voice ('call speaker two Alice').
Meeting Scribe	Records a multi-speaker meeting transcript on-device, tagging you as 'You' and others as Speaker N.
Cocktail Filter (voice focus)	Tries to focus on your voice and reject other speakers.
Voice Guard (biometric + anti-spoof)	Types only when the live speaker matches your enrolled voiceprint and the audio isn't a recording/synthetic.
Glasses↔Desktop Bridge	Dictate from a paired phone/glasses; the desktop does STT + typing.

---

How to turn any of these on

yazses features                 # browse everything with on/off + tier
yazses features enable <name>   # turn one on (installs only its extras)
yazses restart

Experimental features require --force. Advanced ones stay dormant until enabled, so your setup never changes unless you ask.