菜单
Generate short spoken audio files from text with OpenAI text-to-speech. Use when you want a Telegram voice-note reply or any local speech audio from text.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Define Hermit entity schema, templates, and starter records.
Transcribe saved Telegram voice notes and other audio attachments with OpenAI before replying. Use when a Telegram prompt includes an audio file path under inbox/telegram/.
Query the workspace `entities/` tree with helper scripts before manually reading many records.
Define Hermit roles, including `role.md`, `AGENTS.md`, prompts, and startup files.
Maintain the Hermit framework repo separately from workspace data, including framework PR and update flows.
Search `skills.sh` and ClawHub for reusable skills, then install only after user confirmation.
| name | openai-tts |
| description | Generate short spoken audio files from text with OpenAI text-to-speech. Use when you want a Telegram voice-note reply or any local speech audio from text. |
| compatibility | Requires OPENAI_API_KEY or HERMIT_OPENAI_API_KEY, network access to api.openai.com, and a writable output path. |
Use this skill when you want to turn a short reply into a local audio file.
It is especially useful before calling send_telegram_voice_note.
From the repo root:
node skills/openai-tts/synthesize.mjs --text "Short spoken update" --output workspace/inbox/telegram/reply.opus
Useful variants:
node skills/openai-tts/synthesize.mjs --text "I reviewed your plan" --output workspace/inbox/telegram/reply.opus --voice alloy
node skills/openai-tts/synthesize.mjs --text-file workspace/inbox/reply.txt --output workspace/inbox/telegram/reply.opus
node skills/openai-tts/synthesize.mjs --text "Quick update" --output workspace/inbox/telegram/reply.mp3 --format mp3
gpt-4o-mini-ttsalloyopussend_telegram_voice_note.