Skill Selection — DevOps skill for Claude Code
Dynamic skill orchestrator for AI coding agents.
How to install Skill Selection
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open Yuggohel2/skill-selection and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Skill Selection does
Dynamic skill orchestrator for AI coding agents. Scans installed skills, classifies each as infrastructure or project-specific, and locks a minimal manifest — eliminating prompt pollution and skill conflicts. Works on Antigravity, Claude Code, Gemini CLI, Cursor, Windsurf.
Alternatives in DevOps
- Claude Code Router — Use Claude Code as the foundation for coding infrastructure, allowing you to decide how to interact with the m 30.1k ★
- 16 Testing — Prompt 16: Add Test Infrastructure & Smoke Tests 2.3k ★
- Cccc — Coordinate your coding agents like a group chat — read receipts, delivery tracking, and remote ops from your p 1.1k ★
README
skill-selection
**Stop your AI agent from fighting itself.**
A dynamic skill orchestrator that scans your environment, classifies each skill
as infrastructure or project-specific, and locks a minimal manifest — so the agent
uses only what the current project needs.
**Zero hardcoded skill names. Fully dynamic. Works on any platform.**
[Why This Exists](#the-problem) · [Installation](#installation) · [Usage](#usage) · [How It Works](#how-it-works) · [Configuration](#configuration) · [FAQ](#faq)
The Problem
You install 20+ skills to cover design, testing, debugging, mobile, backend, and AI/ML. Then you start a project and your AI agent:
- Applies 5 competing CSS philosophies — brutalist grids mixed with minimalist cards mixed with glassmorphism
- Writes unit tests during a hackathon — burning 3 hours on test scaffolding during a 24-hour prototype sprint
- Loads irrelevant skill instructions — consuming thousands of context window tokens on skills that have nothing to do with the current task
- Produces Frankenstein output — inconsistent code that tries to satisfy every installed skill simultaneously
This is **prompt pollution** — the #1 quality killer in modern agentic coding environments.
How `skill-selection` Differs from Other Orchestrators
Most existing "skill orchestrators" are **task routers** — they break a prompt into subtasks and route Step 1 to Skill A and Step 2 to Skill B. They still leave all 30+ skills active in your environment to collide with each other.
**They orchestrate execution. `skill-selection` orchestrates the environment.**
| Feature | Traditional Task Orchestrators | skill-selection |
|---|---|---|
| Core Question | "Which skill handles step 3?" | "Which skills should even exist in this session?" |
| Competing UI / Style Skills | Left active (causes CSS bleed) | Hard-suppressed (1 skill per domain rule) |
| Infrastructure Protection | Trea |
Related Skills
P2P
Meta-prompt system that writes prompts: raw idea in — a polished, model-specific prompt out. Cross-model (Clau
Agent Devbox
Spin up a disposable cloud dev box with AI coding agents (Claude Code, Codex) pre-installed — one GitHub Actio
Agbx
Run coding agents in isolated, project-specific Docker environments
Rogue Infra Tools Skill
Infrastructure and DevOps tooling (Terraform, K8s, Config Mgmt). Use when the project needs this capability or
Rumpelpod
Orchestrator for LLM coding agents in Docker and Kubernetes pods
Guidance For Dynamic Game Npc Dialogue On AWS
This guidance helps game developers automate the process of creating a non-player character (NPC) for their ga
Related Agents
Stack Detector
Deep project stack discovery that scans all config files, source code, and infrastructure to produce a compreh
SQL
SQL specialist — schema design, indexes, query optimization, migrations, eliminating table scans and N+1 patte
Contract Spec Agent
Emits a DRAFT C# DTO and TypeScript/Angular client interface contract before implementation begins, writing co
