Execute qualquer Skill no Manus
com um clique

Execute qualquer Skill no Manus com um clique

Começar

daw-master

Estrelas1

Forks0

Atualizado23 de maio de 2026 às 20:31

Meta-skill for DAW operations - orchestrates audio transformations and effects pipelines

Instalação

Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.

Executar no Manus

Fonte

sanchorelaxo

sanchorelaxo/hermes-music-sampling

Abrir repositório GitHub Ver repositórios do creator

Download

Executar no Manus

Ocupações relacionadasSOC

Baseado na classificação ocupacional SOC

Técnicos em engenharia de somArtes, Design, Entretenimento, Esportes e Mídia·SOC 27-4014

Explorador de arquivos

63 arquivos

SKILL.md

readonly

name	daw-master
description	Meta-skill for DAW operations - orchestrates audio transformations and effects pipelines
version	0.2.0
author	Hermes Agent
license	MIT
metadata	{"hermes":{"tags":["audio","daw","production","pipeline","transformations"],"related_skills":["daw-master:dawdreamer","daw-master:batch-processor","daw-master:mix-engineer","daw-master:mastering-engineer"]}}

DAW Master

Meta-skill providing a unified interface for DAW operations and audio transformation pipelines.

Purpose

This meta-skill defines standards and base capabilities for audio processing skills under the daw-master namespace. All concrete DAW skills should:

Work with files: Accept input audio file paths, produce output file paths
Be chainable: Operations can be piped/sequenced: sox-normalize -> pitch-shift -> fade-out
Support analysis: Extract metadata about audio (duration, peak, RMS, BPM estimate, etc.)
Expose parameters: Each operation has clear, documented parameters
Return results: Always produce a tangible output (modified file, analysis JSON, etc.)

Philosophy

CLI-first: Skills wrap command-line tools (SoX, FFmpeg, DawDreamer, etc.) but expose a clean Python/Hermes interface.
Stateless operations: Each skill call is independent; no hidden session state.
Composability: Output of one skill feeds into another. Skills are building blocks.
Non-destructive by default: Transformations write to new files; original is preserved.
Extensible: New tools/backends can be added as sub-skills without breaking existing workflow.

Skill Registry

Sub-skills under daw-master:

|| Skill | Tool | Purpose | |-------|------|---------| || dawdreamer | DawDreamer (Python) | Full-featured DAW: effects, mixing, time/pitch, VST | || sox-engine | SoX | Simple edits, effects, conversions | || ffmpeg-audio | FFmpeg | Codec work, filtering, stream ops | || rubber-band-engine | Rubber Band CLI | High-quality time-stretch/pitch-shift | || audio-analyzer | librosa + sonic-annotator | Extract BPM, key, MFCC, loudness | || batch-processor | SoX/FFmpeg scripts | Apply same pipeline to many files | || metadata-manager | BWFMetaEdit + mutagen | BWF/RIFF metadata embedding | || ardour-automator | Ardour (Lua) | Headless session automation & export | || carla-rack | Carla (LV2/VST) | Plugin rack — single-pass multi-effect chains | ||| mix-engineer | SoX/FFmpeg/audio-analyzer | Smart mixing, balance, panning automation | ||| mastering-engineer | FFmpeg/audio-analyzer | Album loudness normalization & QC |

Common Patterns

Pipeline Pattern

# Conceptual: chain skills together
input = "vocals.wav"
pipeline = [
    ("normalize", {"target_level": -6}),
    ("time_stretch", {"factor": 0.95}),  # slow down 5%
    ("reverb", {"room_size": 0.4}),
    ("fade_out", {"duration": 2.0}),
]
output = run_pipeline(input, pipeline)

Discovery Pattern

# Analyze first, then decide transformations
info = analyze_audio("sample.wav")
if info["bpm"] > 140:
    target = "hiphop"
    pipeline = hiphop_chain
else:
    target = "ambient"
    pipeline = ambient_chain

Batch Pattern

# Apply skill to directory
for file in wav_files_in("in/"):
    output = dawdreamer.transform(
        file,
        effects=[gain(-3), compressor(threshold=-20)],
        output=f"out/{stem(file)}_processed.wav"
    )

Mais deste repositório

mesmo repositório

hypno-2

sanchorelaxo/hermes-music-sampling

Use when working with the Sleepy Circuits Hypno 2 video synthesizer/resampler — 2-channel video mixer/looper, shader engine, MIDI CC mapping, CV/modulation, sampler. Firmware v0.0.163+.

2026-06-171

instruments

sanchorelaxo/hermes-music-sampling

Hardware instrument skills — reference guides for each device in the music studio setup.

2026-06-171

hypno2-shader-dev

sanchorelaxo/hermes-music-sampling

GLSL fragment shader techniques adapted for the Hypno 2 video synthesizer — single-pass generative visuals and video effects optimized for Raspberry Pi 5, 5-uniform limit, CC-mapped parameters

2026-05-261

kaoss-dj-mixer

sanchorelaxo/hermes-music-sampling

Korg KAOSS DJ — USB DJ controller with X-Y touchpad, Serato DJ Intro integration, KAOSS effects, EQ, crossfader, and MIDI on Linux

2026-05-261

kaossilator-v1

sanchorelaxo/hermes-music-sampling

Korg KAOSSILATOR dynamic phrase synthesizer (2007 model) — X-Y touchpad, 100 programs, gate arpeggiator, phrase loop recording, USB MIDI on Linux

2026-05-261

kaossilator-v2

sanchorelaxo/hermes-music-sampling

Use when working with the Korg kaossilator 2S dynamic phrase synthesizer — X-Y touchpad, 150 programs, 50 arp patterns, loop recording, audio player, master recorder, microSD storage, and USB MIDI on Linux.

2026-05-261

DAW Master

Meta-skill providing a unified interface for DAW operations and audio transformation pipelines.

Purpose

This meta-skill defines standards and base capabilities for audio processing skills under the daw-master namespace. All concrete DAW skills should:

Work with files: Accept input audio file paths, produce output file paths

Be chainable: Operations can be piped/sequenced: sox-normalize -> pitch-shift -> fade-out

Support analysis: Extract metadata about audio (duration, peak, RMS, BPM estimate, etc.)

Expose parameters: Each operation has clear, documented parameters

Return results: Always produce a tangible output (modified file, analysis JSON, etc.)

Philosophy

CLI-first: Skills wrap command-line tools (SoX, FFmpeg, DawDreamer, etc.) but expose a clean Python/Hermes interface.

Stateless operations: Each skill call is independent; no hidden session state.

Composability: Output of one skill feeds into another. Skills are building blocks.

Non-destructive by default: Transformations write to new files; original is preserved.

Extensible: New tools/backends can be added as sub-skills without breaking existing workflow.

Skill Registry

Sub-skills under daw-master:

Common Patterns

Pipeline Pattern

# Conceptual: chain skills together input = "vocals.wav" pipeline = [ ("normalize", {"target_level": -6}), ("time_stretch", {"factor": 0.95}), # slow down 5% ("reverb", {"room_size": 0.4}), ("fade_out", {"duration": 2.0}), ] output = run_pipeline(input, pipeline)

Discovery Pattern

# Analyze first, then decide transformations info = analyze_audio("sample.wav") if info["bpm"] > 140: target = "hiphop" pipeline = hiphop_chain else: target = "ambient" pipeline = ambient_chain

Batch Pattern

# Apply skill to directory for file in wav_files_in("in/"): output = dawdreamer.transform( file, effects=[gain(-3), compressor(threshold=-20)], output=f"out/{stem(file)}_processed.wav" )