← 返回 Skills 市场
Audio2Text
作者
tinrec-com
· GitHub ↗
· v1.0.0
· MIT-0
194
总下载
1
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install tinrec-audio2text
功能描述
将音频转文字，并提供 AI 总结、录音要点与发言人区分。 音频转文字：云端转写，支持多格式（mp3/wav/m4a 等）。 AI 总结：自动生成摘要与纪要。 录音要点：章节、待办、推荐等结构化要点。 发言人区分：转写结果带发言人标识，便于会议/对话场景。
安全使用建议
This skill is coherent with its stated purpose (cloud transcription): it uploads a local audio file to a Tinrec endpoint, submits a transcription job, polls for completion, and returns transcript, summary, chapters, todos, and speaker data. Before installing/using it consider: 1) Privacy: your audio is uploaded to the Tinrec service (default base URL https://api.tinrec.com/api). Only proceed if you trust that service and its privacy/retention policies. 2) API key handling: the SKILL.md recommends saving user-provided API keys into a local 'api-keys' file; the script also accepts TINREC_API_KEY. Storing secrets on disk can be risky—prefer using short-lived or scoped keys, and avoid leaving keys in plaintext if possible. 3) Manifest mismatch: the skill metadata lists no required env vars, but the code requires an API key; treat that as a minor inconsistency and ensure you supply a valid key via CLI/env when using. 4) Source/trust: the registry entry has no homepage and the owner is an opaque ID — if you need stronger assurance, obtain the script from a known repository or contact the publisher for provenance. If you accept the privacy implications and provide a dedicated API key, the skill appears suitable for its stated use. If you cannot trust the destination service or cannot risk uploading audio, do not use this skill.
功能分析
Type: OpenClaw Skill
Name: tinrec-audio2text
Version: 1.0.0
The skill provides a legitimate utility for transcribing local audio files using the Tinrec API. The Python script (audio2text_cli.py) uses standard libraries to upload audio to the service's storage and retrieve transcription results, with no evidence of malicious execution, data exfiltration beyond the intended purpose, or prompt injection.
能力评估
Purpose & Capability
Name/description (audio→text, summaries, speaker separation) align with the code and SKILL.md: the CLI uploads a local audio file to a Tinrec API, submits a transcription job, polls status, and returns transcript/summary/speaker info. No unexpected third-party services beyond the stated Tinrec endpoints are used.
Instruction Scope
SKILL.md and the CLI instruct the agent/operator to read a local audio file and to persist a user-provided API key in an 'api-keys' file (or provide via env/CLI). The skill will upload the audio and metadata to the Tinrec service (external network transmission) — this is expected for a cloud transcription service, but it is a privacy-sensitive operation and the instructions explicitly require writing/storing API keys locally.
Install Mechanism
No install/spec—instruction-only plus a single Python script that uses only the standard library (urllib/json). Nothing is downloaded or installed by the skill itself; there is no extract or external package install step.
Credentials
Registry metadata lists no required env vars, but the CLI reads an API key from: (1) the --api-key argument, (2) environment variable TINREC_API_KEY, or (3) an api-keys file. The skill therefore effectively requires a service credential, but the manifest did not declare that. This mismatch (undeclared required secret) is an inconsistency to note. The number and scope of environment accesses themselves are otherwise proportionate to the stated purpose (one service API key).
Persistence & Privilege
always:false and no system-wide modifications. The skill asks the operator/agent to save an API key to a local file if desired, but it does not request persistent platform privileges or modify other skills or system configurations.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install tinrec-audio2text - 安装完成后，直接呼叫该 Skill 的名称或使用
/tinrec-audio2text触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
将音频转文字，并提供 AI 总结、录音要点与发言人区分。 音频转文字：云端转写，支持多格式（mp3/wav/m4a 等）。 AI 总结：自动生成摘要与纪要。 录音要点：章节、待办、推荐等结构化要点。 发言人区分：转写结果带发言人标识，便于会议/对话场景。
api-keys，获取地址[https://tinrec.com/api-keys](https://tinrec.com/api-keys)，注册账号后即可免费获取 Key
- 支持多格式音频（如 mp3/wav/m4a）转文字，云端转写速度快
- 新增 AI 自动摘要与音频纪要生成功能
- 增加章节、待办事项、推荐等结构化录音要点输出
- 可区分发言人，会议/对话场景效果更佳
- 优化 CLI 用法，支持传入 API Key 文件或直接传递 Key
- 输出格式包括完整转写结果、要点和发言人信息
Convert audio to text, with AI-powered summaries, key points extraction, and speaker diarization.
**Audio-to-Text:** Cloud-based transcription supporting multiple formats (mp3/wav/m4a, etc.).
**AI Summary:** Automatically generates summaries and meeting notes.
**Key Points:** Structured outputs including chapters, to-dos, and recommendations.
**Speaker Diarization:** Transcriptions include speaker labels, ideal for meetings and conversations.
**API Keys:** Get yours at [https://tinrec.com/api-keys](https://tinrec.com/api-keys) — register an account to receive a free key.
* Supports multi-format audio (e.g., mp3/wav/m4a) for fast cloud transcription
* Adds AI-powered automatic summaries and audio-based meeting notes
* Provides structured outputs such as chapters, action items, and recommendations
* Enables speaker diarization for better clarity in meetings and conversations
* Improves CLI usage with support for API Key files or direct key input
* Output includes full transcripts, key points, and speaker information
元数据
常见问题
Audio2Text 是什么？
将音频转文字，并提供 AI 总结、录音要点与发言人区分。 音频转文字：云端转写，支持多格式（mp3/wav/m4a 等）。 AI 总结：自动生成摘要与纪要。 录音要点：章节、待办、推荐等结构化要点。 发言人区分：转写结果带发言人标识，便于会议/对话场景。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 194 次。
如何安装 Audio2Text？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install tinrec-audio2text」即可一键安装，无需额外配置。
Audio2Text 是免费的吗？
是的，Audio2Text 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Audio2Text 支持哪些平台？
Audio2Text 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Audio2Text？
由 tinrec-com（@tinrec-com）开发并维护，当前版本 v1.0.0。
推荐 Skills