🎙️ dsh-talk

• 1024 store channel: npm i -g dsh1024 once, then dsh1024 plugin --profile web add dsh-talk (counts toward the deepseek1024.com install ranking).

Voice-first session loop for DeepSeek Harness: talk to it, hear it answer.

Press the mic, speak, and the reply is spoken back — with speak-to-interrupt.

English · 简体中文 · Español · Português · हिन्दी

dsh plugin --profile <name> add dsh-talk

---

📖 Ecosystem knowledge base — measured data, not marketing: plugin development guide · plugin-selection data · maintenance criteria.

Install & uninstall

• git channel (latest main): dsh plugin --profile web add github:PerryLink/dsh-talk — the prepare script builds with production dependencies only.
• npm channel (published releases): dsh plugin --profile web add dsh-talk.
• tarball channel: pnpm pack in this repo, then dsh plugin --profile web add ./dsh-talk-<version>.tgz.
• uninstall: dsh plugin --profile web remove dsh-talk (or remove the row from the profile patch).

If pnpm reports ERR_PNPM_IGNORED_BUILDS for this package (esbuild's harmless platform-binary validation), add allowBuilds: { esbuild: true } to your pnpm-workspace.yaml — the dsh CLI prints the exact snippet.

⭐ 如果它帮到了你

这个插件是 DSH 插件家族的一员（95 个，全部 Apache-2.0）。如果你在用，给个 star —— 它不会解锁任何功能，但会让下一个人在搜索里更容易找到它。

English: part of a 95 plugin family for DeepSeek Harness. If it is useful, a star helps the next person find it — nothing is gated behind it.

What is dsh-talk?

Voice-first session loop for DeepSeek Harness: talk to it, hear it answer.

Press the mic, speak, and the reply is spoken back — with speak-to-interrupt.

Workflow overview: install → open panel → act. The examples below use the real command surface.

Compatibility

Surface	Status
Harness	DeepSeek Harness dsh-v0.2.1-alpha.1 (adapted 2026-09-24: third peer clause + engines.dsh + manifestVersion: 1, and the monthly Compat workflow anchored to that line); full gate chain green on 2026-09-24 (dual typecheck rulers, 86 tests, build, self-contained, artifacts, pack). npm dev/test line 0.1.7-rc.2, peers `>=0.1.2-rc.1 <0.2.0
Node	^22.19.0 || >=24.0.0
Browser	Web Speech + MediaRecorder (Chrome/Edge best); host transcription/TTS engines for the rest

What you get

dsh-talk closes the voice loop in both directions:

• speak tool — the agent speaks its replies aloud. TTS engines: the browser voice, edge-tts (network neural voices), or piper (local). Audio plays in the browser; with a local engine it is also written to a file and the tool result carries the absolute path (audioPath) for other tools to consume; on hosts that can carry it, the session log records the sanitized utterance (see Security boundaries).
• Composer mic button — press it, speak, and the transcription lands in the input box (or submits directly). STT engines: the browser's Web Speech (interim results included), a FunASR HTTP server, or local whisper.cpp.
• Speak-to-interrupt — starting to talk stops whatever is playing (client → host over the talk Remote namespace).
• Event announcements — turn completion, pending approvals (waterfall-safe: never blocks the gate), and errors, with a mute switch and configurable phrases.
• Settings tab — engine/language selects and announcement switches, saved as append-only profile-patch operations with backups.

browser                                host
  🎙 press ──▶ interrupt ─────────────────▶ talk/interrupt
  record (MediaRecorder / Web Speech)
  transcribe (browser) or talk/transcribe ─▶ FunASR / whisper.cpp
  setDraft(text) or submit()  ◀── talk:speech projection ── speak tool / announcements
  ▶ play audio (talk/audio or speechSynthesis)

Quick start

dsh plugin --profile web add github:PerryLink/dsh-talk

# 1. install the bundle into your profile
dsh plugin --profile web add github:PerryLink/dsh-talk

# or from npm (published releases)
dsh plugin --profile web add dsh-talk

# 2. restart and verify the row
dsh --profile web --dump-config | grep -A2 'id: talk'

Then press the microphone next to the composer and talk; ask the agent to speak its reply:

> Say "hello" with the speak tool.

Remote browsers need a secure origin

The browser microphone (getUserMedia / Web Speech / MediaRecorder) only works in a secure context: https://… or http://localhost / http://127.0.0.1. A plain http://<LAN-IP>:3081 page — including Safari on another machine — will not expose the microphone, even if the Talk settings are correct.

For a remote browser, use one of:

• open the Web UI through an HTTPS reverse proxy or tunnel;
• SSH-forward the port and open http://localhost:<port> on the client;
• keep the UI on the same machine as the browser.

stt.engine: funasr / whisper changes where transcription runs, not the browser capture permission; recording still needs the secure context.

Examples

These are the real surfaces; the settings tab and mic button are the UI entry points.

Detailed example

Command

speak

Input

{"text":"hello","engine":"browser"}

Output

Spoke hello aloud with the browser engine. No audio file: the browser engine synthesizes in the browser.

Observed from plugins/dsh-talk/tests/lifecycle.spec.ts:124-137; renderer at src/tool.ts:47-64

Failure branch

Input

{"text":"   "}

Output

Error: dsh-talk: speak text must not be empty

Observed from plugins/dsh-talk/src/tool.ts:124; plugins/dsh-talk/tests/tool.spec.ts:41-44

Brief examples

• talk:status
• talk:transcribe

Configuration

All tunables are Schemastery Config fields (changeable from cordis.yml). cordis.patch.yml documents each key inline.

Key	Default	Meaning
record.enabled	true	Show the composer mic button
record.hotkey	(none)	Optional toggle hotkey, e.g. "alt+r"
record.maxSeconds	60	Recording cap in seconds (1..600)
record.autoSubmit	false	Submit the transcription as a user message (false = fill the draft)
record.vad.enabled / silenceMs / energyThreshold	true / 1500 / 0.01	Voice-activity detection: silence auto-ends the recording (degrades when AudioContext is absent)
stt.engine	auto	auto / web / funasr / whisper; auto prefers a configured local engine, then Web Speech
stt.language	auto	BCP-47 language or auto
stt.interim	true	Show interim transcriptions (Web Speech)
stt.silenceFinaliseMs	4000	Stop continuous Web Speech recognition after this many milliseconds without speech (500..15000)
stt.funasr.url	(none)	FunASR inference endpoint; required when the engine is funasr
stt.whisper.modelPath	(none)	whisper.cpp model; required when the engine is whisper
tts.engine	auto	auto / browser / edge-tts / piper; auto prefers piper, then edge-tts, then the browser voice
tts.rate	0	Rate offset in percent (-50..50) for edge-tts/piper
tts.fallbackToBrowser	true	Fall back to the browser voice when a local engine fails
tts.browser.voiceName	(none)	Preferred browser voice name; an unknown name uses the platform default
tts.browser.rate	1	Browser SpeechSynthesis rate (0.1..10)
tts.browser.pitch	1	Browser SpeechSynthesis pitch (0..2)
tts.piper.modelPath	(none)	piper voice model; required when the engine is piper
tts.export.enabled	true	Write locally synthesized audio to a file and return its path as audioPath
tts.export.dir	<home>/.dsh-talk/audio	Export directory; ~ = home, a relative path resolves under home (never the workspace)
tts.export.maxFiles	20	Files retained in the export directory, counting the one just written (1..1000)
tts.export.maxAgeMinutes	1440	Exported files older than this many minutes are pruned on the next write (1..10080)
announce.enabled	true	Master switch for event announcements
announce.onTurnEnd / onApproval / onError	true	Which events are spoken
announce.messages.*	"Turn complete." etc.	Spoken phrases
interrupt	true	Talking stops current playback
maxSpeakChars	20000	Cap on the speak tool's text length (1..100000)
maxAudioCacheBytes	8388608	In-memory synthesized-audio cache cap (1 MiB..64 MiB)

stt.silenceFinaliseMs and record.vad.silenceMs are separate mechanisms: the first finalises the Web Speech transcript when continuous recognition hears no speech, the second is the MediaRecorder energy-based detector that ends the recording (and submits it when record.autoSubmit is on). They run in different pipelines and share no state.

Audio file export

Every utterance synthesized by a local engine (edge-tts, piper) is written to a file in addition to the in-memory cache the browser plays from. The browser engine synthesizes on the client and therefore exports nothing — that state is reported explicitly, never silently.

• Where: tts.export.dir, default <home>/.dsh-talk/audio — the plugin's own directory. It is deliberately never the host's working directory, so a repository is never polluted; a relative tts.export.dir is resolved under the home directory for the same reason.
• Names: speech-<YYYYMMDD>T<HHMMSS>Z-<8 hex>.mp3 (edge-tts) or .wav (piper). The stamp is UTC so names sort chronologically, and the token comes from the utterance id, so a file maps back to its dsh-talk/speech log entry. The spoken text is never part of a file name.
• Cleanup: bounded on every write, age first and count second. Files older than tts.export.maxAgeMinutes (default 1440 = 24 h) are removed, then the oldest files are removed until the directory holds at most tts.export.maxFiles (default 20, counting the file just written). Pruning only ever unlinks names matching this plugin's own speech-… pattern, so other files in a shared directory are never touched. With the defaults the directory holds at most 20 files (~a few MB) and nothing older than a day.
• Failure: exporting is an addition to speech, never a precondition. If the directory cannot be created or written, the utterance still plays and the result reports audioExport: "failed" with the reason; pruning failures are ignored and retried on the next write.
• Model-visible path: the speak tool's canonical value carries audioPath (plus audioExport and, when there is no file, exportNote), and the same path is in the rendered result text — so the agent can pass it straight to another tool that takes a file, for example a messenger plugin's file argument. Event announcements (turn end, approval, error) are synthesized through the same pipeline and are exported the same way; they are bounded by the same caps.
• Handing the file to another plugin: a consumer that only accepts files inside its own workspace — dsh-wechat's send_wechat refuses a file_path outside the calling session's workspace unless its own allowAnyFilePath setting is on — will reject a path under your home directory. For that flow, point tts.export.dir at a directory inside that workspace (for example <workspace>/.dsh-talk-audio); the default stays outside every repository on purpose.

Tools & surfaces

Surface	Kind	Notes
speak	tool	Speaks text aloud (browser/edge-tts/piper); per-call engine/voice overrides; canonical JSON outcome
mic button	conversation.input.left slot	Record → transcribe → fill draft (or submit); speak-to-interrupt on press
settings tab	settings.plugins.tab (id talk)	Engine/language/announcement switches; append-only save
talk:*	Typert Remote	status, audio, transcribe, applySettings, interrupt (host namespace)

Permissions & data

• Permissions: the plugin keeps an in-memory, byte-capped audio cache for playback and, for local engines, writes the synthesized audio to its own bounded directory (tts.export.*, default <home>/.dsh-talk/audio, at most 20 files and 24 h old by default); microphone permission is browser-mediated. The settings tab only appends patch fragments to the profile with a timestamped backup — never rewrites the file.
• Data: audio bytes never enter the model context or the session log. Where the host's session vocabulary accepts it, the dsh-talk/speech event carries the utterance id, engine, reason, size, sanitized text, and browser voice/rate/pitch when applicable; on envelope-less hosts the event is not written at all. The speak tool result carries the exported file's path (a persistent artifact in the plugin's own directory), and everything else on a display/log surface passes the redaction helpers, so credentials, JWTs, bearer headers, and temp paths stay out.
• Network: only the engines you configure are contacted. edge-tts performs network synthesis, FunASR uses its configured endpoint, and Chrome's webkitSpeechRecognition sends microphone audio to Google's servers for transcription; browser speechSynthesis playback remains local.

Security boundaries

• Model-visible ⟺ logged — the model sees only the speak tool's canonical value and render text. The dsh-talk/speech event is appended only when the host can carry it (see Host compatibility); the tool/call + tool/result events always remain the reconstructable trail.
• Approval announcements never block — the approval/request listener always calls next().
• Sanitized output — credentials and temp audio paths never reach logs or displays. The one deliberate exception is the exported audio path in the speak result, which is a persistent file in the plugin's own directory (still passed through the secret redactor) and is meant to be handed to another tool.
• Host compatibility — the dsh-talk/speech event is appended through a yes/no gate. Hosts whose known-type vocabulary covers the event append it, and the call reports that it did; every other host — every released line through 0.1.1-rc.2, the 0.1.2-alpha line, 0.1.2-rc.1, and also 0.1.6-alpha.2, whose Session.append can stamp only the surface intent and not the ignorable envelope (that field is retained for stored-log read compatibility only) — gets no append, so speech can never pollute the session log there. The skip is not silent any more: the host keeps per-session appended/skipped counters and talk/latest(sessionId) reads them back, the client's session-scoped playback list stays empty, and the speak tool results remain the reconstructable audit trail.
• Fail loud — invalid engines, out-of-range values, and engines configured without their required model/endpoint fail the mount.

Known limitations

• Browser support: Web Speech and MediaRecorder are feature-detected; without them the mic button disables itself and host engines (FunASR/whisper.cpp) can still transcribe when configured.
• Local engines are your install: edge-tts, piper, and whisper.cpp executables and models must be installed separately.
• Audio files need a local engine: the browser voice synthesizes on the client, so speak returns no audioPath there (the result says so explicitly), and a local engine that fails and falls back to the browser also has no file. Exported files live in the plugin's own bounded directory — see Audio file export for the location, naming, and cleanup rules.
• Recording format: the browser records with its native MediaRecorder codec; whisper.cpp may require a WAV-configured recorder or a server-side conversion for other formats.
• Settings apply on reload: the settings tab appends to the profile patch; a profile reload (or web-app restart) activates the changes.
• Live playback history is empty on hosts without the vocabulary: on 0.1.1-rc.2, the 0.1.2-alpha line, 0.1.2-rc.1, and 0.1.6-alpha.2 the host vocabulary does not know dsh-talk/speech, so the gate writes nothing and the client's session-scoped playback list stays empty; talk/latest(sessionId) still answers with the newest utterance and the skipped counter. Speech itself, the mic, the settings tab, and the tool are unaffected.
• Legacy logs written by dsh-talk ≤ 0.2.1 may need repair before cold load: versions through 0.2.1 appended unmarked dsh-talk/speech events. On hosts 0.1.0-rc.7 and newer, a session whose log already contains them fails its next cold load with SessionFormatUnsupportedError. Repair: stop the host, back up the session's .jsonl log, add "ignorable":true as a top-level member of every JSON line whose "type" is "dsh-talk/speech" (for example, insert "ignorable":true, right after the opening {), then reopen the session. Nothing else changes and nothing is lost; new appends from this version never add unmarked events.

Development

pnpm install        # node ^22.19 || >=24
pnpm run typecheck  # tsc: src + tests against the local harness checkout
pnpm run typecheck:ci  # tsc against the published 0.1.7-alpha.2 types (no paths)
pnpm test           # vitest: 108 tests, 16 suites
pnpm run build      # tsc declarations + tsdown bundles (lib/)
pnpm run verify:self-contained  # dependency specs resolve from the registry
pnpm run verify:artifacts       # built ESM faces + client ModuleLoader handshake
pnpm pack           # the published tarball

Interoperability with other DSH plugins

Verified against DSH 0.2.0-rc.2 (the runtime this README ships for) and the high-star plugin set surveyed on 2026-10-05.

This plugin does not interfere with other plugins, including the widely installed high-star ones:

• No tool-name collision. Every tool is namespaced; no bare name owned by a shipped tool or another plugin is registered.
• No service-key collision. It provides talk; that key is not a built-in seam and is not provided by any surveyed high-star plugin.
• No slot collision. It registers no client slot key, so it cannot contend for a shadows-shipped-ui seat.
• No HTTP route collision. It registers no webServer prefix.
• No patch-layer collision. The bundle patch only inserts its own row; it never overrides a built-in row's config.
• No global mutation. It does not patch prototypes, rewrite process.env, or replace the global fetch dispatcher.

Shared event listeners are non-interfering by construction. It observes the ordering-sensitive event approval/request with ctx.on() — Cordis's broadcast registration, where every listener runs and none can starve another. Every listener here delegates through next(), so the chain is never short-circuited, and a mutation is applied to the value next() produced rather than returned in its place:

Static evidence: dsh-plugin-doctor K10–K13 report pass for every check on this repository.

Topics

dsh, dsh-plugin, deepseek-harness, deepseek, cordis, voice, speech, tts, stt, speech-to-text, text-to-speech, microphone

Contributors

• @Mo-Mia — Web Speech privacy/compatibility, composer input actions, language fallback, continuous dictation, talk:speech projection, and browser TTS design proposals (#2–#9); several informed the current settings surface.
• @PerryLink — creator and maintainer: speak pipeline, speech engines, mic recorder, event announcements, projection unit, and the five-language docs.

PerryLink DSH Plugin Family

This project is part of the DeepSeek Harness plugin family from PerryLink. The roster stays collapsed below so every README keeps a short first screen; expand it to see every row and its status.

Expand all plugins

Plugin	Status
dsh-aqua-input-check
dsh-archive-check
dsh-auto-review
dsh-autotier
dsh-bid-ca-precheck
dsh-bid-qual-check
dsh-cert-mcp
dsh-checkpoint-rewind
dsh-click
dsh-composer-history
dsh-contract-stance
dsh-customs-doc-check
dsh-data-quality
dsh-demurrage-ledger
dsh-doublecheck
dsh-drill-script-check
dsh-eia-guide-check
dsh-emergency-plan
dsh-essay-rubric-check
dsh-evidence-check
dsh-export-ctl-check
dsh-fast
dsh-fmea-table-check
dsh-forecast-penalty
dsh-fund-research
dsh-github
dsh-gongwen-flow-check
dsh-gongwen-word-check
dsh-guard-plan-qc
dsh-hazchem-check
dsh-hazplan-check
dsh-hidden-risk-map
dsh-hs-classify
dsh-icd-rule-check
dsh-industry-research
dsh-lawcite-adapter
dsh-laya
dsh-lc-doc-check
dsh-learning-gap-check
dsh-library
dsh-local-ai
dsh-lsp-actions
dsh-mask
dsh-mcp-panel
dsh-medrec-qc
dsh-memento
dsh-nsfc-form-check
dsh-nurse-record-check
dsh-observe
dsh-origin-rvc-check
dsh-ota-review-check
dsh-output-styles
dsh-paper-doc-adapter
dsh-permission-rules
dsh-permit-report-check
dsh-pipeline-check
dsh-pleading-draft
dsh-plugin-doctor
dsh-plugin-guide
dsh-plugin-kit
dsh-plugin-upgrade
dsh-policy-brief-draft
dsh-power-loss-split
dsh-power-ticket-check
dsh-ppap-check
dsh-protest-deadline
dsh-railway-window
dsh-repair-order-qc
dsh-research-report
dsh-review-reply-check
dsh-rulefile-check
dsh-safety-brief-check
dsh-score
dsh-session-sync
dsh-site-log-check
dsh-skill-pack-security
dsh-soe-decision-check
dsh-soilwater-check
dsh-sop-sync-check
dsh-spc-gbt-adapter
dsh-talk
dsh-tender-extract
dsh-tender-matrix
dsh-test-drive
dsh-ticktick
dsh-translate
dsh-warranty-calc
dsh-budget	🧊 FROZEN — see the repo README
dsh-claude-move	🧊 FROZEN — see the repo README
dsh-defend	🧊 FROZEN — see the repo README
dsh-draw	🧊 FROZEN — see the repo README
dsh-reach	🧊 FROZEN — see the repo README
dsh-background-agents	🚫 RETIRED — see the repo README
dsh-session-pin	🚫 RETIRED — see the repo README
dsh-team-rooms	🚫 RETIRED — see the repo README

License

Apache License 2.0 © 2026 dsh-talk contributors