first-principles

first-principles is a skill for Claude Code, Codex from tupe12334/instinct. It costs 21 tokens per session (2,087 once invoked), scanned A, original, MIT.

A way to solve problems by separating verified facts from assumptions, then building a solution from those basic facts.

In plain words
What is it for?
Use it to examine difficult decisions, question constraints, and develop solutions when existing examples are not enough.
Why use it?
It helps avoid copying common approaches that may not fit the actual problem.

Skill for Claude CodeCodex

Part of the instinct plugin — 54 skills, 1 hook shipped together

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.

agentmods
npx agentmods add skills/tupe12334/instinct/first-principles
Any agent
npx skills add tupe12334/instinct --skill first-principles
Clone the repo
git clone --depth 1 https://github.com/tupe12334/instinct

Made for: Claude Code, Codex.

Or install instinct, the plugin that ships this one along with the rest of its 54 skills, 1 hook.

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 first-principles

README.md
[![agentmods](https://agentmods.dev/badge/skills/tupe12334/instinct/first-principles.svg)](https://agentmods.dev/skills/tupe12334/instinct/first-principles)
Your own site
<a href="https://agentmods.dev/skills/tupe12334/instinct/first-principles"><img src="https://agentmods.dev/badge/skills/tupe12334/instinct/first-principles.svg" alt="Measured on agentmods" height="20"></a>
Per session 21 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,087 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00021 $0.02087
Opus 5 $0.00010 $0.01043
Sonnet 5 $0.00004 $0.00417
Haiku 4.5 $0.00002 $0.00209

Measured 3d ago against content hash c3de84beb829, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

first-principles 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 3d 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/first-principles/SKILL.md · 137 lines

How it starts

The opening of the file, as written. The whole thing — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.

First-Principles Thinking

Overview

First-principles thinking is a reasoning method that decomposes a problem to its most basic, verified truths and reconstructs a solution from there — rather than reasoning by analogy from what others have done. It was practiced by Aristotle and is closely associated with how physicists approach novel problems.

Conventional thinking:        First-principles thinking:
  "Others do X, so we do X"     Problem
         ↓                         ↓
    Analogy / copy            Ask: what is fundamentally true here?
         ↓                         ↓
   Incremental tweak          Strip assumptions → bedrock facts
                                   ↓
                              Rebuild solution from facts up
                                   ↓
                              Novel or superior solution

The key discipline is distinguishing what is actually true from what is conventionally assumed.

Core Concepts

Assumptions vs. Fundamental Truths

An assumption is any constraint accepted without verification: cost, time, materials, method. A fundamental truth is a fact that would hold regardless of convention — physics, math, verified data, human need. First-principles thinking hunts for assumptions masquerading as truths.

Decomposition

Breaking a problem into its component parts until you reach irreducible elements. You cannot decompose "battery is expensive" further by analogy; you can decompose it into: cathode material cost, anode material cost, electrolyte cost, manufacturing overhead — each of which can be addressed independently.

Reconstruction

Once you hold only verified facts, you rebuild a solution unconstrained by how it has been done before. The reconstruction step is creative and deliberate — the facts constrain what is possible, not what should be built.

The Socratic Test

For each constraint or belief, ask: "How do I know this is true?" If the answer is "everyone does it this way" or "we've always done it this way," it is an assumption, not a truth. Valid answers cite data, physical laws, or direct measurement.

Read the full file on GitHub · 137 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. 3d ago First seen · 137 lines · 21 tokens per session scan A c3de84beb829

Subscribe to this mod's changes

first-principles is a skill published in the GitHub repository tupe12334/instinct (1 stars, last pushed 17d ago), licensed MIT. It adds 21 tokens to every session and 2,087 once invoked, about $0.0001 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 2026-08-31.