---
name: skill-law
description: Apply the miscsubjects Laws of Skills when deciding whether something becomes a skill, before creating or editing any skill, when judging a skill edit, when a failure or owner correction repeats, or when importing an external skill.
---

# Apply the Laws of Skills

This Skill is the model-operating expression of [the human article](/a/skill-law). Read the canonical object at /api/articles/skill-law when exact clauses or provenance are needed.

## The axiom

A skill is the build's memory of a failure it refuses to repeat. Unwritten memory repeats; exhibit-less memory is fiction; unjudged edits decay into noise. Fluency is not provenance — one model's confident output never silently becomes another model's instruction.

## Decide

- Did a model actually fail at this without the skill, and where is the exhibit?
- Is this a procedure (skill) or a fact (instruction file) — and could a mechanical gate enforce it instead?
- Does an existing skill already cover this concern — edit it rather than create a sibling?
- Who wrote this rule, and what proof binds it — a failing exhibit, or the owner's recorded word?

## Operate

1. Search both skill trees for the concern before writing anything new.
2. Capture the failing exhibit: the ledger event, owner correction, or wrong output.
3. If a mechanical gate can block the failure, build the gate; keep prose for judgment only.
4. Write or edit the smallest clause that prevents the failure; one concern per skill.
5. Sync the sibling tree in the same change; claim shared files before editing.
6. Judge the edit: fresh agent, baseline-fails then with-skill-passes; link both exhibits.
7. Record actor, exhibit, and delta in the ledger; the skill points at the receipt.
8. Project the skill to its site object so every model reads the same law from one identity.

## Reject as nonconforming

- skill written from imagination with no failing exhibit
- correction acknowledged in chat but never written into the skill
- edit landing in one tree only
- description that summarizes the procedure instead of stating the trigger
- skill weakened to excuse a past output
- model-authored rule binding others without proof or acceptance
- duplicate skills accreting on one concern
- prose restating what a validator already blocks

## Pair with

shared-failure-to-skill (the capture procedure), shared-rule-capture (owner restatements), shared-write-law (claim before edit), writing-skills (authoring mechanics), skill-creator (eval harness), test-driven-development and systematic-debugging (the code-writing discipline this law governs the upkeep of).

## Return

Return only: the exhibit, the clause delta, the gate (if mechanical), the fresh-agent verdict, the ledger receipt, both trees synced. Cite kao:skill-law, its version, and the exhibits used.