← 返回 Skills 市场
Oasis Audio Pro
作者
littlelightsai
· GitHub ↗
· v1.0.1
· MIT-0
86
总下载
1
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install oasis-audio-pro
功能描述
Oasis Audio Pro is an AI audio narration generator that transforms your current state of mind, content you want to digest, or recent life events into a perso...
安全使用建议
This skill appears to do what it says: it reads local agent/session files to personalize audio, composes a single prompt, and sends that prompt to eagle-api.xplai.ai using a visitor JWT. Before installing or using it, consider: 1) The skill will read conversation history and USER.md under ~/.qclaw and ~/.openclaw — only proceed if you are comfortable with that local read access. 2) It persists a visitor JWT (~/.oasis_audio/oasis_config.json), a consent record, and an audit.log that may contain prompt text; inspect those files and set restrictive filesystem permissions if needed. 3) The code includes a sensitive-content preview and requires explicit confirmation before sending prompts flagged by heuristics — verify that flow in practice (run with --dry-run to test). 4) If you keep highly sensitive data in your local sessions, test the skill first with non-sensitive content and review the audit.log and consent behavior. If you want extra assurance, review the full xplai_gen_audio.py and context_collector.py flow (search for where persistent consent is enforced and where write_audit_log is called) or run the scripts in a restricted environment.
功能分析
Type: OpenClaw Skill
Name: oasis-audio-pro
Version: 1.0.1
Oasis Audio Pro is a personalized audio generation skill that accesses local OpenClaw/QClaw chat history and user profiles to tailor content. While this involves high-privilege access to sensitive local data, the implementation demonstrates strong defensive intent and transparency. Key security features include a mandatory local consent mechanism, extensive regex-based redaction of secrets (API keys, SSH keys) and PII (medical, financial, and legal details) in xplai_gen_audio.py, and a 'sensitive preview' requirement that halts execution for user confirmation if sensitive content is detected. Data sent to the external API (eagle-api.xplai.ai) is limited to the final composed prompt, with raw history and session files processed strictly locally.
能力标签
能力评估
Purpose & Capability
Name and description claim an AI audio generator using xplai.ai; included scripts obtain a visitor JWT, compose a prompt, and call eagle-api.xplai.ai. Reading ~/.qclaw and ~/.openclaw for personalization is coherent with the stated purpose. No unrelated credentials or binaries are requested.
Instruction Scope
SKILL.md and the code explicitly instruct/implement reading session .jsonl, memory .md, and USER.md files under ~/.qclaw and ~/.openclaw for personalization (read-only). The skill claims raw session files are not transmitted; the code composes a distilled prompt and sends only that. This is within scope, but it does access potentially sensitive local conversation history — the consent and sensitive-preview steps are important guardrails.
Install Mechanism
No external install/downloads or package fetching are declared; this is an instruction-only skill with bundled Python scripts. That minimizes install-time risk. No URLs or archives are fetched at install.
Credentials
The skill requests no environment variables and uses a visitor JWT (no user API key) — proportional to the function. It persists a JWT (~/.oasis_audio/oasis_config.json) and writes an audit.log and consent file in the script directory; those persisted files may store sensitive prompt text locally unless properly redacted/confirmed.
Persistence & Privilege
always:false (normal). The skill persists a visitor token, a consent record (.oasis_audio_consent.json), and an audit.log in the skill's directory. Persisting local artifacts is reasonable for a client tool, but users should be aware these files exist and review their contents/permissions.
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 oasis-audio-pro - 安装完成后，直接呼叫该 Skill 的名称或使用
/oasis-audio-pro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
v1.0.1
Oasis Audio Pro 1.0.1
- Updated description and documentation to clarify that local Python scripts read-only process session/memory files in ~/.qclaw or ~/.openclaw for personalization.
- No user API key required; a visitor JWT token is obtained automatically and stored in ~/.oasis_audio/oasis_config.json on first use.
- Confirmed that only the final composed prompt is sent to xplai.ai—never raw conversation history or user files.
- No changes to code or functional behavior.
v1.0.0
Oasis Audio Pro 1.0.0 – Initial Release
- Generate personalized AI audio narration from user notes, content, or recent life events with optional background music.
- Supports English & Chinese, 9 reference narration modes, and custom profiles for unique scenarios.
- Local context analysis: only composes and sends prompts, never raw history or sensitive files, to the xplai.ai API.
- Built-in sensitive info protection: shows a sanitized prompt preview and requires explicit user confirmation before sending sensitive content.
- Automated result polling and easy, zero-config setup.
元数据
常见问题
Oasis Audio Pro 是什么？
Oasis Audio Pro is an AI audio narration generator that transforms your current state of mind, content you want to digest, or recent life events into a perso... 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 Oasis Audio Pro？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install oasis-audio-pro」即可一键安装，无需额外配置。
Oasis Audio Pro 是免费的吗？
是的，Oasis Audio Pro 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Oasis Audio Pro 支持哪些平台？
Oasis Audio Pro 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 Oasis Audio Pro？
由 littlelightsai（@littlelightsai）开发并维护，当前版本 v1.0.1。
推荐 Skills