메뉴
Convert text to speech with Google Gemini TTS API at fixed 1.2x playback speed and WAV output. Use when users need fast narration generation from English or Chinese text, with optional voice and output filename.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
| name | gemini-tts-fast |
| description | Convert text to speech with Google Gemini TTS API at fixed 1.2x playback speed and WAV output. Use when users need fast narration generation from English or Chinese text, with optional voice and output filename. |
Convert text to speech using Google Gemini's TTS API with fixed 1.2x playback speed. Automatically generates standard WAV audio files optimized for video narration.
google-genai packageGOOGLE_API_KEY environment variable (stored in .env file)When the user requests text-to-speech conversion, the skill will automatically apply 1.2x speed:
.env file to get the API key--output=filename.wav (optional, default: output.wav)--voice=VoiceName (optional, default: Puck)The skill now includes its own tts_cli.py script in the skill directory, making it fully portable:
set -a && source .env && set +a && \
source venv/bin/activate && \
python .claude/skills/gemini-tts-fast/tts_cli.py "<text>" --output="<filename>" --voice="<voice>" --speed=1.2
Note: Users need to:
pip install google-genai.env file with GOOGLE_API_KEY=your-keybrew install ffmpeg (macOS) or equivalentpython -m venv venvGOOGLE_API_KEY is missing from .env, instruct user to add itbrew install ffmpeg)Simple usage (English):
/gemini-tts-fast "Hello world"
→ Generates output.wav at 1.2x speed
With custom output:
/gemini-tts-fast "Welcome to our app" --output=welcome.wav
→ Generates welcome.wav at 1.2x speed
With custom voice:
/gemini-tts-fast "Thank you for listening" --output=thanks.wav --voice=Aoede
→ Generates thanks.wav at 1.2x speed with Aoede voice
Chinese text:
/gemini-tts-fast "你好世界" --output=hello_cn.wav --voice=Kore
→ Generates hello_cn.wav at 1.2x speed with Kore voice
Processing script.json:
/gemini-tts-fast @script.json
→ Automatically processes all narration scenes from script.json at 1.2x speed
atempo filter1.2x speed is optimal for video narration because:
Turn fragmented knowledge, rough notes, research snippets, interview takeaways, project learnings, paper excerpts, or partially formed ideas into structured archive cards for the Personal Blog archive system. Use this whenever the user asks to remember, 整理, 沉淀, 归档, 记到 archive, turn notes into cards, merge new knowledge into an existing archive note, or when they provide scattered points that should become a reusable archive card rather than a polished blog post. Also use this when new input should be matched against existing archive/blog content so similar notes are merged instead of duplicated.
Use when the user explicitly asks for a desktop or system screenshot (full screen, specific app or window, or a pixel region), or when tool-specific capture capabilities are unavailable and an OS-level capture is needed.
Turn paper links, blog posts, articles, or any worthy external content into curated entries for the Personal Blog curated section. Use this whenever the user sends a link and says 精选/curated/收录/加到这个模块/收录到精选/加到精选, or when you discover notable external content worth surfacing on the homepage. Each curated entry is a concise recommendation with your own reading notes — not just a link dump.
Explore a website that exposes a well-known agent manifest (RFC-8615 path) and answer questions or summarize content. Works on any site whose owner ships /.well-known/<name>-manifest.json with a structured tree, a public content endpoint dictionary, and natural-language instructions for agents. Use when the user asks "what's on <site>?", "summarize the latest post on <site>", "find resources about X on <site>", or wants you to act as a reader/agent on someone's personal site or content service.
Automatically engage with X (Twitter) home timeline by liking and replying to relevant tweets. The agent reads a persona profile (X-PROFILE.md), fetches timeline via script, filters tweets by relevance, generates comments directly, and publishes via script. Designed to run as a scheduled cron job (3x daily). Use when the user asks to set up X auto-engagement, configure a commenter persona, run timeline engagement, or troubleshoot engagement failures.
Convert written articles into structured video scripts with timing, narration, and scene breakdowns