Rewrite vague instruction documents into clear specifications with a change report.
npx skills add https://github.com/keithmackay/plsfix --skill plsfix
Please help me install this Agent Skill. Skill: plsfix Source: https://github.com/keithmackay/plsfix/tree/main Command: npx skills add https://github.com/keithmackay/plsfix --skill plsfix
Converts vague, ambiguous, or poorly structured instruction documents into clear, actionable specifications so humans and AI can execute them reliably. It diagnoses structural and content failures, preserves author voice, and surfaces assumptions for confirmation.
Improve this instruction document for clarity and produce a rewritten version plus a change report.
High-intent search queries and answers about installing and using this skill.
To turn vague product requirements into clear specs, this Skill rewrites ambiguous instruction documents into actionable specifications and generates a change report mapping every edit to one of twelve clarity principles.
Your system prompt or SKILL.md gets misinterpreted by AI due to missing output contracts, structural ambiguity, or contradictory instructions. This Skill diagnoses these failures using a 12-principle audit and rewrites the document for reliable execution.
To clarify a runbook or brief without inventing scope, this Skill preserves the author's voice, makes minimal edits, and marks inferred details with [CONFIRM] so you can validate assumptions before execution.
Yes, you can rewrite a prompt template and track changes. This Skill produces a rewritten document alongside a detailed change report that maps each structural and content edit to the specific principle it implements.
The best way to audit reusable prompt templates for structural and content failures is to run a 12-principle diagnostic checklist that locates ambiguity, contradictions, and missing delivery contracts before rewriting.
Yes, improving instruction documents requires defining output contracts and examples. This Skill ensures rewritten specifications include these delivery details so both humans and AI can execute the instructions reliably.
toobulkehRun a skill on a test input, self-grade, and rewrite with A/B decisions.
TomGranotTriage bounced contacts flagged by email_health_flag for deletion or recovery.
gaoqiongxieVersion AI coding prompts in Git for deterministic, testable code generation.
skeletorfletClarify complex requirements through structured Socratic questioning.
