What specifically does Claims Validator look for?

It targets percentage-based metrics, time/cost estimates, comparative claims without baselines, and mentions of CLI commands that aren't actually implemented in the code.

Can it tell the difference between a real benchmark and a fake one?

It flags any metric that doesn't include a citation or link to a methodology, prompting the user to either provide evidence or remove the claim.

How does it know if a feature is implemented?

The skill analyzes the codebase to verify if the commands or flags described in your documentation actually exist as functional code.

Does this skill work on non-Markdown files?

While optimized for Markdown documentation like READMEs, it can analyze any text-based content or code comments for unverifiable statements.

Claims Validator

Name: Claims Validator
Author: jmagly

byjmagly

•

学習とドキュメンテーション

Validates technical documentation by identifying unsupported metrics, unverifiable claims, and marketing superlatives.

概要

Claims Validator is a specialized utility for Claude Code designed to audit documentation, README files, and technical guides for accuracy and integrity. It systematically identifies 'technical fluff' such as performance metrics without benchmarks, cost estimates without data, and comparative statements lacking baselines. By cross-referencing feature descriptions with the actual codebase, it ensures that your documentation only promises what the software actually delivers, helping maintain developer trust and project transparency.

主な機能

Identifies cost and time estimates that lack supporting usage data
Flags subjective marketing superlatives like 'revolutionary' or 'game-changing'
67 GitHub stars
Scans for unsubstantiated percentage improvements and performance benchmarks
Detects mentions of commands or features that do not exist in the codebase
Generates detailed reports with specific line-by-line recommendations

ユースケース

Auditing a project README for accuracy before a major version release
Ensuring internal documentation reflects the current state of implemented features
Reviewing technical marketing materials for unsubstantiated performance claims

概要

主な機能

Identifies cost and time estimates that lack supporting usage data
Flags subjective marketing superlatives like 'revolutionary' or 'game-changing'
67 GitHub stars
Scans for unsubstantiated percentage improvements and performance benchmarks
Detects mentions of commands or features that do not exist in the codebase
Generates detailed reports with specific line-by-line recommendations

ユースケース

Auditing a project README for accuracy before a major version release
Ensuring internal documentation reflects the current state of implemented features
Reviewing technical marketing materials for unsubstantiated performance claims