moai-workflow-docs-claim-check

Validates documentation claims against supplied evidence and labels each atomic claim.

1.2k|222|Updated Sep 16, 2025
One-click install
npx skills add https://github.com/modu-ai/moai-adk --skill moai-workflow-docs-claim-check
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: moai-workflow-docs-claim-check
Source: https://github.com/modu-ai/moai-adk/tree/main/internal/template/templates/.claude/skills/moai-workflow-docs-claim-check
Command: npx skills add https://github.com/modu-ai/moai-adk --skill moai-workflow-docs-claim-check

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Public-facing documents like READMEs and release notes often contain claims that are unsupported, outdated, or unverifiable, and manually auditing them claim by claim is slow and error-prone.

Core Features & Use Cases

  • Atomic Claim Triage: Extracts every checkable statement from a document and splits composite claims so each carries exactly one assertion.
  • Evidence-Based Labeling: Walks an ordered decision tree to label each claim as verified, stale-suspected, unsupported, or needs-human, with named evidence anchors.
  • Strict Read-Only Boundaries: Runs no commands, produces no fixes, and performs no code or security review, certifying this in every report.
  • Use Case: Before publishing a release note, supply the note plus your manifest and CI logs to receive a three-section report showing which claims are verified, which are stale, and which need a human to run a specific command.

Quick Start

Ask the assistant to check whether the claims in your README are supported by the evidence files you provide, such as your package manifest and CI logs.

Frequently Asked Questions about moai-workflow-docs-claim-check

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check if my README claims are accurate?

Supply the README along with evidence files such as manifests, CI logs, or release listings. The skill extracts every checkable claim, splits composite statements into atomic ones, and labels each as verified, stale-suspected, unsupported, or needs-human.

What documents can be audited for unsupported claims?

Public-facing documents are in scope: READMEs, release notes, install and usage guides, quickstarts, and migration notes. Internal design notes, task trackers, and private runbooks are explicitly out of scope.

Does the claim check run tests or commands to verify claims?

No. The skill never executes commands, builds, or tests. When a claim can only be settled by running something, it names the exact command and target file and labels the claim needs-human.

What is the difference between stale-suspected and unsupported claims?

Stale-suspected means the claim was once true but evidence shows a version, date, count, or name has drifted. Unsupported means the evidence never carried the claim, with a reason of missing-evidence, contradicted, or insufficient-coverage.

Can the claim check fix incorrect documentation it finds?

No. The skill is read-only and produces findings only. It describes what a maintainer would change and where, but never emits patches, diffs, or rewritten passages.