Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add chuancyzhang/deepseek-cowork --skill speech-to-textgit clone --depth 1 https://github.com/chuancyzhang/deepseek-coworkWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/chuancyzhang/deepseek-cowork/speech-to-text)<a href="https://agentmods.dev/skills/chuancyzhang/deepseek-cowork/speech-to-text"><img src="https://agentmods.dev/badge/skills/chuancyzhang/deepseek-cowork/speech-to-text/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/chuancyzhang/deepseek-cowork/speech-to-text"><img src="https://agentmods.dev/badge/skills/chuancyzhang/deepseek-cowork/speech-to-text.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00035 | $0.01441 |
| Opus 5.5 | $0.00014 | $0.00576 |
| Sonnet 5.5 | $0.00007 | $0.00288 |
| Haiku 4.5 | $0.00003 | $0.00144 |
Grade A, and why
speech-to-text scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.
语音转文字
按能力设置使用本地 SenseVoice 组件或 OpenAI 兼容 /audio/transcriptions 接口转录音频或视频。配置由用户在能力设置中管理,不要向用户索取或在 Tool 参数中传递 API Key、接口地址和远程模型名。
必须先确认输出隐私模式
在调用转录工具前,必须先询问用户是否需要 AI 润色,并清楚说明两种模式:
- 不润色:转录正文只写入工作区,工具结果不会把正文返回给当前对话模型。当前模型只能看到路径和运行状态等元数据,不得再读取生成的转录文件。本地模式不上传音频;远程模式仍会把音频发送到用户配置的 ASR 接口。
- AI 润色:转录完成后,原始文字会交给当前会话正在运行的模型处理;如果当前模型是云端服务,转录文字会发送给该服务商。润色结果随后写入工作区。
如果用户已经明确选择,不要重复询问。
工作流
- 确认是否 AI 润色。若本轮恰好有一个由界面保留的语音转文字专用附件,调用工具时省略
audio_path,避免把附件路径写进模型上下文。 - 调用
transcribe_audio,polish必须与用户选择一致。通常省略model、diarize和speaker_count,由已保存的后端配置决定默认行为。 - 当
polish=false:- 工具已将原始稿写入工作区。
- 只向用户报告输出路径,并根据工具返回的
backend准确说明音频由本地组件处理或已发送到远程接口,同时说明正文没有提供给当前对话模型。 - 不得调用任何文件读取工具打开该转录稿,也不得声称自己看过正文。
- 当
polish=true:- 使用工具返回的
transcript在当前模型中润色。 - 保留原文已有的时间戳、Speaker 标签、原意和事实;只修正标点、断句、口头语和高置信度识别错误,不总结、不翻译、不合并不同说话人。
- 工具会先把原文作为可恢复的最终稿写入
output_path;调用save_transcript_result,以polished=true安全替换为润色稿。 - 如果无法可靠完成润色,直接展示工具返回的原始稿;原文最终稿已经存在,也可调用
save_transcript_result,以polished=false明确确认原文结果。
- 使用工具返回的
- 最终告诉用户输出文件路径、转录后端、是否经过 AI 润色、识别模型,以及工具实际返回的说话人数和时长。
后端差异
- 本地模式使用 SenseVoice,默认执行说话人分离和时间戳对齐,并支持中文、英文、日文、韩文和粤语。
- OpenAI 兼容接口模式上传原始音频并采用响应 JSON 的非空
text字段,不调用本地语音组件,不自动切换或回退到本地模型。 - 远程模式不提供本地说话人分离或时间戳;不要传入
model,不要设置diarize=true或正数speaker_count。用户需要说话人标签时,明确建议切换到本地模式。
本地说话人分离规则
- 默认自动估算说话人数;用户知道人数时,将其作为
speaker_count传入。 - 输出使用
Speaker 1、Speaker 2等中性标签,不猜测真实姓名或身份。 - 同一时间检测到多人说话时保留组合标签,例如
Speaker 1 + Speaker 2。 - 无法可靠对齐到任何说话人的文字标记为
Speaker ?,不得凭上下文猜测。 - 说话人分离或时间戳对齐失败时明确报错,不退化成伪造的单说话人结果。
本地模式首次使用
本地语音模型和运行依赖统一由“设置 → 组件与依赖 → 语音转文字组件”管理。用户从 Cowork GitHub Release 获取独立的 Windows x64 语音组件 ZIP,再点击“选择安装包”离线部署。远程模式不需要该组件。
如果组件未安装或损坏,transcribe_audio 会明确返回 component_not_ready,提醒用户选择 Release 安装包进行安装或修复;不得在 Tool 调用期间或组件安装期间静默下载模型、运行 npm 或切换下载源。组件就绪后只使用本地文件。
边界
- 只处理用户明确指定的工作区文件或本会话附件。
- 当本会话已指定本 Skill 时,界面会把音频附件保留给该 Skill:当前对话模型只看到不含文件名和路径的占位说明,工具从本轮附件授权元数据中选择文件。
- 支持 WAV、MP3、M4A、AAC、FLAC、OGG、Opus、MP4 和 WebM;不支持实时麦克风听写。
- 本地模式下,长音频会切成短片段识别,再按绝对时间戳与整段说话人分离结果对齐;处理时长会随录音长度增加。远程模式把单个原始文件直接交给接口,大小与时长限制由该服务决定。
- 本地模式下,如果 FFmpeg 检测到损坏或不可解码的音频包,结果必须明确展示
warnings,并说明转录只覆盖可解码区间。 - 默认不覆盖已有输出文件;需要覆盖时必须得到用户明确同意并传入
overwrite=true。 - 不自动生成摘要、翻译、会议纪要或字幕文件,除非用户另行要求。
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 66 lines · 35 tokens per session scan A 1ce90bcbe5d7
speech-to-text is a skill published in the GitHub repository chuancyzhang/deepseek-cowork (18 stars, last pushed today), licensed Apache-2.0. It adds 35 tokens to every session and 1,441 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-04.
Other skills, from other repositories
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…