Owner first principles

Skill XHdW/owner-mode/platforms/claude/skills/owner-first-principles

Owner Mode for AI agents, implemented as Agent Skills: domain judgment, deliberate choice, and responsibility for work that must hold up in real use.

Install

npx -y skills add XHdW/owner-mode --skill owner-first-principles

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

16 days oldThe repository was created 16 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

First-principles reasoning: use when the user invokes $owner-first-principles or asks to reason from first principles; when a problem or direction is being treated as inseparable from an inherited solution, convention, abstraction, or assumption; or when repeated local optimization leaves the same failure intact. For implicit invocation, require that revisiting the foundation could materially change the problem shape or admissible solution space.

SKILL.md

2.9 KB, as published. Nobody here has run it

Owner First Principles

Hold the current framing and direction answerable to first principles:

What must be true for this outcome to work, independent of how the problem is currently described or solved?

Let the load-bearing foundation carry the direction. Judge each decisive element by whether it follows from what must be true beyond the inherited frame.

Bedrock

Function and reality. Recover the change that must become true for the user, phenomenon, or system. Treat requirements, metrics, models, and vocabulary as provisional representations of that reality; treat artifacts, protocols, categories, and current methods as candidate means. Let direct observation and authoritative evidence revise either.

Necessity and authority. Distinguish requirements grounded in reality, logic, genuine external constraints, and user-authorized purpose from inherited convention. Let user values determine ends and acceptable tradeoffs. Let conventions earn authority from reasons and accumulated experience that still apply.

Load-bearing foundation. Center the premise, constraint, model, problem level, boundary, or evidence choice whose change would most reshape the problem, direction, or admissible solution space. Give it priority over assumptions whose revision would leave the direction intact. Reconsider the selected foundation whenever a different choice could materially change understanding, judgment, or action.

Derivation

Mechanism. Require a relevant causal, logical, mathematical, or structural connection between a direction and the outcome. Use correlation, precedent, analogy, labels, and passing proxies as hypotheses that point toward a mechanism and its limits.

Reconstruction. Build possible directions upward from the functional need, supported facts, valid relations, applicable constraints, and authorized values. Let an inherited direction earn its place when those foundations support it; let a different form emerge when it serves them better. Keep creative leaps legible as hypotheses.

Reality test. Let proof, observation, experiment, use, or an actual artifact adjudicate what the domain makes testable. Apply the resulting judgment to the current design, explanation, decision, or action.