midi-cleanup

midi-cleanup is a skill for Claude Code, Codex from glincker/ableton-skills. It costs 76 tokens per session (1,175 once invoked), scanned A, original, MIT.

A MIDI-editing workflow for polishing recorded or programmed notes without changing the musical idea. MIDI is data describing notes, timing, velocity, and other performance instructions rather than recorded audio.

In plain words
What is it for?
Use it to humanize velocity and timing, quantize notes, improve voice leading, clean duplicate or stuck notes, and refine MIDI clips.
Why use it?
It fixes mechanical timing, awkward chord movement, duplicate or stuck notes, and uneven playing while preserving the part's intended character.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it to humanize velocity and timing, quantize notes, improve voice leading, clean duplicate or stuck notes, and refine MIDI clips.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/glincker/ableton-skills/midi-cleanup
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add glincker/ableton-skills --skill midi-cleanup
Clone the repo
git clone --depth 1 https://github.com/glincker/ableton-skills

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for midi-cleanup

README.md
[![agentmods](https://agentmods.dev/badge/skills/glincker/ableton-skills/midi-cleanup/github.svg)](https://agentmods.dev/skills/glincker/ableton-skills/midi-cleanup)
Your own site
<a href="https://agentmods.dev/skills/glincker/ableton-skills/midi-cleanup"><img src="https://agentmods.dev/badge/skills/glincker/ableton-skills/midi-cleanup/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for midi-cleanup

Your own site · 80×15
<a href="https://agentmods.dev/skills/glincker/ableton-skills/midi-cleanup"><img src="https://agentmods.dev/badge/skills/glincker/ableton-skills/midi-cleanup.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 76 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,175 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00076 $0.01175
Opus 5 $0.00038 $0.00588
Sonnet 5 $0.00015 $0.00235
Haiku 4.5 $0.00008 $0.00118

Measured 10d ago against content hash ffafaaa87c92, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

midi-cleanup scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/midi-cleanup/SKILL.md · 103 lines

How it starts

The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.

MIDI Cleanup

You are polishing a MIDI clip the user wrote or recorded. Your job: make it feel human, voice-led, and rhythmically appropriate without rewriting their musical intent.

Workflow

1. Identify the target clip

User must specify which clip. If ambiguous, ask:

"Which clip — the lead on track 3 or the chords on track 5?"

Then read the clip with get_notes_from_clip (or equivalent). Note the note count, range, and rhythmic density.

2. Identify the cleanup type

Match the user's request to one of these operations:

A. Humanize velocity

For drums, hats, percussion, or any rhythmic part that feels stiff.

  • Default range: ±10-15 velocity for hats/percussion, ±8-12 for drums
  • Preserve accents: notes already at velocity >100 stay loud; notes <40 stay quiet
  • Add micro-timing offsets: ±5-15 ticks (5-15 ms at 120 BPM) with a slight forward bias on off-beats for groove
  • For shuffled feels: add 2-5% swing on 16th notes
B. Humanize timing

For piano, guitar, or string parts that sound MIDI-stiff.

  • ±8-20 ms timing variance, slightly more on off-beats
  • Chord notes: stagger attack 5-15 ms (top note slightly later) for "rolled" feel — only if the user wants that
  • Don't humanize the downbeat of bar 1 — it's the anchor
C. Voice leading fix

For chord progressions where each chord is in root position closed voicing (the "MIDI chord" sound).

Rules:

  • Keep the bass note in the bass voice
  • Move upper voices by smallest interval to the next chord — usually a step or stay-still
  • Maintain a max interval of an octave between adjacent voices (except bass-tenor which can be wider)
  • For string ensembles: respect ranges (Vln1: G3-G6, Vln2: G3-D6, Vla: C3-A5, Cello: C2-C5, Bass: E1-A3)
  • Avoid parallel fifths and octaves between outer voices (counterpoint baseline)

Apply by reading the current chord notes, computing the optimal voicing for each chord, and rewriting.

D. Quantize (intelligent)

Don't auto-quantize 100% — that destroys feel.

Read the full file on GitHub · 103 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 10d ago First seen · 103 lines · 76 tokens per session scan A ffafaaa87c92

Subscribe to this mod's changes

midi-cleanup is a skill published in the GitHub repository glincker/ableton-skills (24 stars, last pushed 17d ago), licensed MIT. It adds 76 tokens to every session and 1,175 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

ableton-live

Connect and operate the Loophole Bridge (Ableton Live MCP server). Use when the user wants to check Live/bridge prerequisites, wire an MCP client to Loophole, or run a Live editing recipe (humanize, build arrangement, batch rename, chord from prompt). Triggers: "is my Ableton bridge working", "set up loophole"…

OthmanAdi/loophole · 90 tokens

livepilot-creative-director

Use when the user makes an open-ended creative production request — "make it feel like X", "sound like X", "develop this", "mutate", "more interesting", "less generic", "take it somewhere", structure decisions without exact specs, or any reference/style ask. Also use when routerequest returns…

dreamrec/LivePilot · 103 tokens

livepilot-core

Core discipline for LivePilot — agentic production system for Ableton Live 12. 474 tools across 57 domains. This skill should be used whenever working with Ableton Live through MCP tools. Provides golden rules, tool speed tiers, error handling protocol, and pointers to domain and engine skills.

dreamrec/LivePilot · 64 tokens

livepilot-corpus-builder

Build a private knowledge corpus from your own Live projects, rack library, Max devices, and plugin presets. Use when the user says "build my corpus", "scan my projects", "index my racks", "make atlas of my library", "private atlas from my files", or wants Claude to know their personal sound (not just Ableton's…

dreamrec/LivePilot · 78 tokens

livepilot-sound-design-engine

This skill should be used when the user asks to "design a sound", "analyze a patch", "fix a static sound", "add modulation", "check my timbre", "improve a synth patch", or wants critic-driven sound design feedback and iterative patch refinement.

dreamrec/LivePilot · 63 tokens

livepilot-devices

This skill should be used when the user asks to "load a device", "add an effect", "find a plugin", "device chain", "rack", "preset", "sound design setup", "load instrument", "find a synth", or wants to browse, load, and configure devices in Ableton Live.

dreamrec/LivePilot · 69 tokens