Loading…
当使用者说"让 AI 写谱不要直接生成音频"、"做一个 AI 作曲的后端"、"LLM 输出 MIDI 怎么做"、"生成的音符老是越界/被截断"、"要能导出 .mid 给 DAW"时使用。用 LLM（BYOK / OpenAI 兼容接口）从零生成可用的管弦乐 MIDI 与分轨音频：两级调用（整曲和声蓝图 + 逐乐器音符表）、提示词结构、被 max_tokens 截断时的残稿抢救、确定性校验修复、渲染与 MIDI 导出，以及全程的规则作曲降级策略。降级事实必须如实写进元数据，不假装是模型写的。参考实现是 MaestrWave 的 M7 写谱演奏模式。
llm-midi-composition is an agent skill entry in ChatHome Skills, sourced from GitHub Skill Search and scored 58/100. # LLM 生成 MIDI 的完整方法 从「调音频模型要一段 WAV」换成「AI 先写谱、采样器再演奏」时，产物仍是每件乐器一条音轨，下游（浏览/播放/混音/指挥）无感知，同时谱子落盘可导出 MIDI。核心是一条流水线： ``` 蓝图（LLM 写和声走向）→ 分谱（LLM 每件乐器写音符）→ 校验修复 → 渲染成音轨. As an agent skill, it packages a reusable instruction set that AI coding agents such as Codex and Claude Code can load to perform this task without hand-written glue code. ChatHome Skills discovered this entry through its daily ingest of official registries and GitHub discovery queries, and re-verifies repository facts, freshness signals, and install metadata on every run. The quality score weighs official-registry presence, a working install command, license clarity, GitHub star trajectory, and repository freshness — the full formula is on the Sources page. The install command above is ready to copy. Review the last-pushed date and repository status before adding llm-midi-composition to a production agent workflow; the snapshot facts below are point-in-time values from the most recent ingest.
codex skill install ChromiteCr/SkillsReview the upstream source before running this command.
Scan coverage: Install metadata and the indexed document body.
Automated signals from install metadata and a bounded skill body — not a full code security audit and not a recursive source or dependency review.
# LLM 生成 MIDI 的完整方法 从「调音频模型要一段 WAV」换成「AI 先写谱、采样器再演奏」时，产物仍是每件乐器一条音轨，下游（浏览/播放/混音/指挥）无感知，同时谱子落盘可导出 MIDI。核心是一条流水线： ``` 蓝图（LLM 写和声走向）→ 分谱（LLM 每件乐器写音符）→ 校验修复 → 渲染成音轨 → 落盘 + 导出 .mid ``` #