Before AI — AI skill for Claude Code
Before you reach for AI — interrogate one real workflow from first principles, delete the dead weight, then decide if AI is even warranted.
How to install Before AI
Installs to ~/.claude/skills/saifalyounan-before-ai-before-ai/SKILL.md
mkdir -p ~/.claude/skills/saifalyounan-before-ai-before-ai && curl -fsSL https://raw.githubusercontent.com/SaifAlYounan/before-ai/HEAD/commands/before-ai.md -o ~/.claude/skills/saifalyounan-before-ai-before-ai/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Before AI does
description: Before you reach for AI — interrogate one real workflow from first principles, delete the dead weight, then decide if AI is even warranted. Asks; never advises. argument-hint: [the workflow that annoys you, e.g. "approving a new subsidiary CEO"]
You are running `/before-ai`, a structured, Socratic walkthrough for a knowledge worker — usually a lawyer, **not** an engineer. Your job is to take ONE real workflow they bring and help them decide, in this order: what to **delete*
Alternatives in AI
- DontFeedTheAI — Transparent anonymization proxy for AI-assisted pentesting 654 ★
- Ooda Subagents — OODA Loop AI Agents - Observe, Orient, Decide, Act framework for systematic problem-solving 170 ★
- Deadcode Reaper Skill — 扫出并清除「自称已废」的墓碑 skill、AI 幽灵碎片与死代码,统计释放的 Token 与磁盘 46 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Canny
Stops AI coding agents from claiming work is done without evidence. Deterministic hooks decide, TypeSafe's Jev
V Review Plan
Run an independent cross-model (Codex) adversarial review of a Compound V plan/manifest before dispatch, then
Headsign
A tiny phase gate for coding agents: your shell checks decide pass/fail (never the LLM), your workflow routes
Assess Agentcore Migration
Assess an EKS-hosted agent against the Agentic AI Lens; logs the walk, then proposes changes for the customer
LLM Wiki Check Updates
Check origin repo for newer framework files (read-only; determines the origin from FRAMEWORK-SYNC.md, then Git
Stepwarden
Every tool call your agent makes, checked before it runs. A Claude Code plugin that uses TypeSafe AI's Jev to
Related Agents
Ponytail Review
Cold over-engineering reviewer. Spawned by parallel workflows when non-test source changes, to find what to de
Panel Mechanics
Review panel seat four. Catches naming, dead code, convention drift, and comments that should not exist. Read-
Dead Code Finder
Use to find unused exports, unreachable branches, and never-imported files in a codebase. Reports findings wit