guard-artifact-scopelisted

Prevent internal constraints from leaking into reader-facing artifacts or becoming unsupported product claims. Use if named. Invoke implicitly only when one artifact mixes product content with engineering, workflow, editorial, verification, or history constraints and a concrete unresolved scope conflict remains. Artifact type alone is not a trigger; skip routine phase-owned edits.
railgun20001/the-first · ★ 0 · AI & Automation · score 70

Install: claude install-skill railgun20001/the-first

# Guard Artifact Scope Constraints may govern an artifact without becoming content in that artifact. Apply this skill quietly and keep the current workflow moving. ## Invoke only for a real conflict Explicit invocation always applies. Otherwise invoke only when both conditions hold: - One artifact spans reader-facing product content and an internal engineering, workflow, editorial, verification, or history constraint. - An unresolved risk exists of misplacing that constraint or claiming an outcome without evidence. Do not invoke it merely because the task edits a README, UI string, PRD, technical document, release note, or handoff. Continue without it when the phase skill already identifies each statement's owner and proof. Do not create a new phase, gate, status, or confirmation step. Ask only when the conflict would materially change the artifact or product. ## Establish the artifact contract 1. Identify the readers and the decision or action the artifact supports. 2. Identify its authoritative domain and reuse that domain's source of truth. 3. Separate product behavior and user boundaries from internal constraints. 4. Include a statement only when it changes a reader's action, is needed to use the product safely, communicates a real capability or limitation, or belongs in this artifact's authority. Otherwise omit it or keep it in the correct technical, workflow, verification, or decision source. Do not create a document just to record the classification. Preserve