Podcast Generator

播客生成器 — 根据用户描述，通过搜索引擎抓取最新资讯，生成口语化播客脚本，根据脚本语义自动匹配最合适的讯飞TTS声音，合成时长3分钟内的MP3音频并发送。触发词：生成播客、播客、podcast、帮我做一段音频、做一期节目。

Before installing or enabling: (1) confirm the skill author/source — there is no homepage and metadata omits required env vars; (2) only provide XFYUN_APP_ID / XFYUN_API_KEY / XFYUN_API_SECRET if you trust the skill and consider using a limited-scope or disposable XFYUN key; (3) verify that the xfyun-tts skill referenced is one you installed from a trusted source and that its TTS script (tts.py) is what will be executed; (4) be aware the skill will fetch external web pages (scraping arbitrary search results) — do not ask it to summarize or fetch private or sensitive topics; (5) note it uses find ~/.openclaw to locate tts.py which reads your OpenClaw skill directory — if you have sensitive files there, consider moving them or running the skill in a sandbox; (6) if you require stronger assurance, request the publisher to update registry metadata to declare the XFYUN env vars explicitly and provide a verifiable source/homepage or supply an install spec so you can audit the referenced xfyun-tts implementation.

Type: OpenClaw Skill Name: podcast-gen Version: 1.0.0 The skill is a podcast generator that automates news retrieval via search engines and audio synthesis using the xfyun-tts skill. It is classified as suspicious because it instructs the AI agent to construct and execute shell commands, including searching the local filesystem (~/.openclaw) and running a Python script with dynamically generated arguments. While the use of quoted heredocs in SKILL.md suggests some awareness of shell safety, the pattern of executing shell commands based on AI-selected parameters (like the voice ID) represents a high-risk capability that could be exploited via prompt injection.