← 返回 Skills 市场
86
总下载
1
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install nex-voice
功能描述
Voice note transcription and intelligent action item extraction for capture and organization of verbal communication. Record and transcribe voice notes, voic...
安全使用建议
This package appears to do what it says: local transcription with optional cloud LLM processing. Key things to consider before installing:
- Privacy: By default audio and transcripts stay in ~/.nex-voice. Only enable the LLM features if you are comfortable sending transcript text to the configured API provider and you trust that provider. The API key can be stored in ~/.nex-voice/config.json or provided via AI_API_KEY; keep that secret safe.
- Explicit opt-in for external calls: The code only transmits data when you (a) configure an API key and (b) use the --use-llm option (or set LLM configuration). If you never configure an API key and avoid the --use-llm flag, transcripts remain local.
- Implementation notes: setup.sh attempts to run lib/storage.py --init to initialize the DB, but storage.py does not provide a command-line --init handler; this is likely a harmless bug (you can initialize by running the CLI commands or creating the DB through the Database class). The package also expects the whisper CLI and ffmpeg to be installed separately (setup.sh warns about them but does not install them). Review the config file (~/.nex-voice/config.json) before enabling LLM features.
- Recommendation: If you plan to use LLM extraction, review lib/action_extractor.py (the LLM call) to confirm which endpoint you will contact and how the prompt is constructed, and store the API key in a safe place. If you need strict offline operation, do not configure an API key and avoid --use-llm. Otherwise the skill is internally coherent and matches its advertised purpose.
功能分析
Type: OpenClaw Skill
Name: nex-voice
Version: 1.0.0
The nex-voice skill is a legitimate utility for local voice note transcription and action item extraction using OpenAI Whisper and FFmpeg. It stores all data locally in the ~/.nex-voice directory and only performs external network requests (via curl in lib/action_extractor.py) if the user explicitly configures an LLM API key and provider. The code uses safe subprocess execution patterns, lacks any evidence of data exfiltration or persistence mechanisms, and the SKILL.md instructions are strictly aligned with the tool's stated purpose.
能力标签
能力评估
Purpose & Capability
Name/description match the code and CLI: the package transcribes audio using Whisper, extracts action items, stores data under ~/.nex-voice, and can optionally call an external LLM. Required binaries (python3, whisper, ffmpeg) are appropriate for the stated purpose. The presence of code files (transcriber, storage, action extraction) aligns with functionality.
Instruction Scope
SKILL.md instructs only the expected operations (transcribe, extract actions, search, manage items). The only runtime behavior that leaves the machine is the optional LLM extraction: when --use-llm is requested and an API key is configured the code issues an HTTP request (via curl) to the configured API base with the transcript in the prompt. Otherwise audio and transcripts are kept locally under ~/.nex-voice. This external call is explicitly optional and documented, so behavior is within scope but you should be aware transcripts may be transmitted if you enable LLM features.
Install Mechanism
There is no remote download/install spec in the registry entry. The bundle includes a setup.sh that is an idempotent local setup script which only creates directories and checks for system tools (it does not fetch or execute remote archives). That is low-risk. Note: setup.sh invokes python on lib/storage.py with a --init flag which does not appear to be handled by that module (likely a minor bug/logic error, not a network or supply-chain action).
Credentials
Registry metadata declares no required env vars. The code uses optional AI_API_KEY / AI_API_BASE / AI_MODEL (via config) only for the optional LLM pathway. Requesting an API key only makes sense for optional LLM features and is proportionate; however if you configure an API key (stored in ~/.nex-voice/config.json or in AI_API_KEY env) transcripts will be sent to that provider when you opt in. No unrelated credentials are requested.
Persistence & Privilege
The skill is not forced-always and does not request elevated privileges. It stores data and config files under the user's home (~/.nex-voice) and does not modify other skills or system-wide agent settings. Autonomous invocation is enabled (default) but that is normal and is not combined with broad credential access in this package.
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 nex-voice - 安装完成后，直接呼叫该 Skill 的名称或使用
/nex-voice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
Initial release
元数据
常见问题
Nex Voice 是什么？
Voice note transcription and intelligent action item extraction for capture and organization of verbal communication. Record and transcribe voice notes, voic... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 86 次。
如何安装 Nex Voice？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install nex-voice」即可一键安装，无需额外配置。
Nex Voice 是免费的吗？
是的，Nex Voice 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Nex Voice 支持哪些平台？
Nex Voice 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 Nex Voice？
由 Nex AI（@nexaiguy）开发并维护，当前版本 v1.0.0。
推荐 Skills