Skip to content

Latest commit

 

History

7 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Alp Discipline

A decision discipline for the moment you meet a boundary.

What it is

The Alp Discipline has a single job: to help a human or an AI agent decide better at the moment they meet a boundary. It is not a project management method, a craft guide, or a compliance standard. It mandates no ceremonies, no tools, and no checklists, and it moves no decision authority from owners to executors. It teaches the one thing a competent practitioner with good intentions cannot derive alone, and leaves the rest to their judgment.

Who it is for

Anyone, in any field, sector, or role: an engineer, a researcher, an operations coordinator, a solo founder, or an AI agent. It requires no technical background.

How to read it (5 minutes)

File What it is Time
ESSENCE.md The single compressed surface: four principles, five questions, four floors 1 min
DISCIPLINE.md The canon: purpose, seven definitions, four principles, and the discipline's own non-goals 5 to 8 min
APPLYING.md Optional mechanics for those who want them; nothing in it is mandatory as needed
EXAMPLES.md Worked examples across fields: research, operations, journalism, software as needed

The four principles at a glance

(This is an introductory summary. The single authoritative compressed surface is ESSENCE.md.)

  1. Draw the negative space first. Name what you will not do, and prove that you understand the load-bearing boundaries by imagining their violation.
  2. Extension is granted to necessity, not to approach. A boundary that blocks your approach is doing its job. Only exhausted alternatives prove a necessity, and you never grant yourself the extension.
  3. An honest stand on a route that leads to loss. Carry verified evidence and the smallest counterproposal to the decision owner. Silent execution is never acceptable.
  4. The right altitude, a durable step. Declare whether the work is a patch, a slice (a bounded, coherent change), a design, or constitution work. Every step advances durably, and papering over is legitimate at no altitude.

For AI agents

Give your agent ESSENCE.md, whole and unmodified. ESSENCE is the agent's single canonical Alp Discipline surface, and it operates within every higher priority constraint the agent already has: system and safety policies, the law, and explicitly granted authority. The five questions tell the agent how to think; the four floors tell it what never to do. Keep exactly one compressed surface, because parallel variants drift apart.

Implementations

The discipline stays single and simple; its implementations may be unlimited. Integrations, agent prompt templates, schemas, and enterprise adoption models belong to separate repositories, not to this one. This repository names no tool, no product, and no organization, because the discipline must remain adaptable to any environment.

Status

v1.0.4. The core, meaning DISCIPLINE.md and ESSENCE.md, was frozen on 2026-07-18; the remaining files are packaging and change freely. The core reopens only on evidence from real usage, meaning a concrete situation in which the text gave no answer, gave a wrong answer, or was systematically misread; that is the discipline's own Principle 2 applied to itself. Every change, core or packaging, is recorded in AMENDMENTS.md. If you have such evidence, open an issue.

Translations

Turkish, the source language in which v1.0 was authored and frozen, lives in translations/tr/. The root English text is canonical for publication. Any divergence between the two is a defect, not a fork; please open an issue.

License and attribution

MIT. Created by Alperen Sartacoglu. Embedding ESSENCE.md in a machine-consumed prompt requires no attribution display.

About

A decision discipline for the moment a human or an AI agent meets a boundary. Four principles, five questions, four floors.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors