Airsc is free for everyone. If it helps you, support it with $2–$20.

Support
PluginCautionAgent workflow & memorySome setup

slopware-skills

transcendr · transcendr/slopware-skills

This collection provides portable AI agent skills and plugins, including the MSW Kernel for Minimum Sufficient Work, to optimize agent workflows.

Free, portable AI agent skills and plugins for Codex, Claude Code, and Agent Skills clients by Slopware Engineer (@aienginerd). Home of the MSW Kernel for Minimum Sufficient Work.

36 stars2License: No licenseUpdated 25 days ago0 opens

Preview

▍

What you can do with it

  • Implement the MSW Kernel to ensure agents only admit and complete necessary work with proof.
  • Use the CODER Loop plugin to optimize Codex development through bounded work and fresh evaluation.
  • Apply the Timebox skill to converge authorized work within a defined time and budget.

Safety check

Caution

Usable, but read the notes first.

  • ●No clear license: check with the author before using it in a commercial product.

Automated scan of the repo files. Always review code you run.

From the README

Slopware Skills

Free forever ♡

Portable, individually installable skills and plugins for AI coding agents by Slopware Engineer.

Packages

PackageTypePurpose
MSW KernelSkill and pluginAdmit only necessary work, require proof, and stop at the fixed point.
MSW HookOptional pluginReinforce the MSW Kernel once at session context boundaries.
MSL KernelSkill and pluginReport only what the reader must act on, in a form they can check.
TimeboxSkill and pluginConverge authorized work inside AWT and a closeout-only CGP.
CODER LoopCodex skill and pluginOptimize Codex development through bounded work, fresh evaluation, and evidence-scoped remediation.
Codex Voice OptimizerCodex skill and pluginOptimize Codex Voice for hands-free orchestration across owning work threads.

Each package has its own documentation and remains independently installable.

The Slopware Dev Stack

A modular Codex-native stack that removes waste at every development boundary. Install one layer or compose the stack. Free forever.

ComponentStack roleContribution
Codex Voice OptimizerControl planeHands-free intent, routing, authority, and material progress
CODER LoopDevelopment engineBounded implementation, independent evaluation, remediation, acceptance, and evolving model routing
MSW KernelScope kernelNecessary coordination, work, findings, repairs, and proof
TimeboxConvergence envelopeOne explicitly authorized AWT/CGP clock around the work
MSL KernelCommunication kernelThe final facts the user must understand or act on

The product story is simple: CVO controls it. CODER builds it. MSW keeps it necessary. Timebox makes an authorized run converge. MSL makes the result clear. Every package works alone; together they form an evidence-evolving, high-efficiency agentic development system without a hard dependency or shared runtime.


MSW Kernel: Minimum Sufficient Work

All necessary work. Nothing beyond it.

Package documentation · Skill source

Minimum Sufficient Work is the principle. The MSW Kernel is the compact instruction set and program that applies the principle to agent work.

The MSW Kernel is a lightweight alternative to heavier, more invasive agent-control systems. Its small instruction surface can be just as effective, and arguably more effective, because it gives the agent one necessity test, a proof obligation, and a clear stopping condition without layering on another workflow.

MSW also protects contract integrity: temporary development state must neither add product responsibilities nor weaken required behavior. Necessary environment work stays at its own scope, and current capability and proof gaps must be reported truthfully.

It applies to implementation, debugging, review, research, planning, and documentation, not just code size. The original MSW Kernel post introduced the idea.

Install the MSW Kernel

Codex

This uses a free-form Git marketplace directly from this repository. It does not require submission to or verification for an official marketplace.

codex plugin marketplace add transcendr/slopware-skills
codex plugin add msw@slopware-skills

Add automatic reinforcement only if you want it:

codex plugin add msw-hook@slopware-skills

Start Codex, open /hooks, and trust the single SessionStart hook. The msw plugin remains hook-free.

Claude Code

claude plugin marketplace add transcendr/slopware-skills
claude plugin install msw@slopware-skills

Add automatic reinforcement only if you want it:

claude plugin install msw-hook@slopware-skills

The msw plugin remains hook-free.

skills.sh

This installs the portable, hook-free Agent Skills version:

npx skills add transcendr/slopware-skills --skill msw -g

The installer supports Codex, Claude Code, and other Agent Skills-compatible clients.

Generic ~/.agents/skills

git clone https://github.com/transcendr/slopware-skills.git
mkdir -p ~/.agents/skills
cp -R slopware-skills/plugins/msw/skills/msw ~/.agents/skills/msw

Optional hook

The separate msw-hook plugin injects one compact MSW reminder when a session context starts, resumes, clears, or compacts. It does not run before tools, after edits, or on every user prompt. There are no modes, daemons, dependencies, status lines, or state files.

Install only msw when you want explicit invocation with no hook.

FAQ

How is this different from heavier solutions like Ponytail?

Ponytail and similar systems can introduce modes, decision ladders, persistent reinforcement, invasive hooks, and highly opinionated conventions. The MSW Kernel stays at a lower level: define the contract, test necessity, prove the required work, and halt. This avoids overthinking, process doom spirals, and conventions that can clash with your own while remaining applicable beyond coding tasks.

What makes this work?

The contract converts an open-ended request into the smallest provable outcome. The deletion test forces every proposed action to justify itself against that outcome instead of habit, reviewer authority, or imagined future needs. The proof obligation prevents minimalism from becoming corner-cutting. The fixed point gives the agent a concrete stopping condition, so it neither stops early nor kee

View source on GitHub