How do I install the Fact Check skill?

Run npx skillmds@latest add carlheath/fact-check in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Fact Check safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Fact Check?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Fact Check free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Fact Check?

carlheath (@carlheath) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Fact Check by carlheath — Agent Skill

Name: Fact Check
Author: carlheath

Fact Check

Fact-checking and quality review of texts. [WHAT] Verifies claims, checks sources, identifies style issues. Can be used standalone or piped after /chronicle. [WHEN] Use when: fact-check, verify, check, quality check. [EXPERTISE] Fact-checking, source verification, style review.

carlheath 6c65830 2.4 KB Updated Sep 17, 2026

File contents

Review: $ARGUMENTS

Fact-Check — verification and quality review

Purpose

Review text for:

Factual errors — verifies claims against sources
Source quality — assesses credibility
Style issues — identifies AI patterns or style breaks
Reference format — checks APA 7

Workflow

STEP 1: extract claims

List all verifiable factual claims in the text:

Statistics and numbers
Quotes and attributions
Technical descriptions
Historical claims
Dates

STEP 2: verify each claim

For every claim:

Search for confirmation from credible sources
Mark as: ✅ VERIFIED | ⚠️ PARTIAL | ❌ WRONG | ❓ NOT VERIFIABLE
If WRONG: propose correction with source

STEP 3: assess source quality

Review the reference list:

Are the sources credible? (academic, established media, primary sources)
Are they current? (post-2022 for AI-related material)
Is APA 7 format correct?
Are full URLs included?

STEP 4: style review

Check for common AI patterns:

"Not just X but also Y"
"It's important to note..."
"In summary..."
Overuse of short fragments as paragraphs
Monotone sentence length

STEP 5: summary

Produce the review report:

## Review Report

**Text:** [Title]
**Date:** [YYYY-MM-DD]

### Fact-check
| Claim | Status | Comment |
|-------|--------|---------|
| ... | ✅/⚠️/❌ | ... |

### Corrections needed
1. [Specific correction with source]

### Source quality
- [Assessment]

### Style observations
- [Any issues]

### Recommendation
[ ] APPROVED for publication
[ ] CORRECTIONS REQUIRED (see above)

Use in pipe

When /fact-check runs after /chronicle:

Read the produced text
Run the full review process
If APPROVED: continue to next step
If CORRECTIONS: list them and await feedback

Delegation

For extensive fact-checking, delegate to a research agent:

"Verify the following claims from a text:
[list of claims]

MAX: 5 web searches
RETURN: verification status for each claim with source"

🎯 COMPLETED: [SKILL:fact-check] [Fact-check completed] 🗣️ CUSTOM COMPLETED: [SKILL:fact-check] [Review done]

---name: fact-checkdescription: Fact-checking and quality review of texts. [WHAT] Verifies claims, checks sources, identifies style issues. Can be used standalone or piped after /chronicle. [WHEN] Use when: fact-check, verify, check, quality check. [EXPERTISE] Fact-checking, source verification, style review.---Review: $ARGUMENTS# Fact-Check — verification and quality review## PurposeReview text for:1. **Factual errors** — verifies claims against sources2. **Source quality** — assesses credibility3. **Style issues** — identifies AI patterns or style breaks4. **Reference format** — checks APA 7## Workflow### STEP 1: extract claimsList all verifiable factual claims in the text:- Statistics and numbers- Quotes and attributions- Technical descriptions- Historical claims- Dates### STEP 2: verify each claimFor every claim:1. Search for confirmation from credible sources2. Mark as: ✅ VERIFIED | ⚠️ PARTIAL | ❌ WRONG | ❓ NOT VERIFIABLE3. If WRONG: propose correction with source### STEP 3: assess source qualityReview the reference list:- Are the sources credible? (academic, established media, primary sources)- Are they current? (post-2022 for AI-related material)- Is APA 7 format correct?- Are full URLs included?### STEP 4: style reviewCheck for common AI patterns:- "Not just X but also Y"- "It's important to note..."- "In summary..."- Overuse of short fragments as paragraphs- Monotone sentence length### STEP 5: summaryProduce the review report:```## Review Report**Text:** [Title]**Date:** [YYYY-MM-DD]### Fact-check| Claim | Status | Comment ||-------|--------|---------|| ... | ✅/⚠️/❌ | ... |### Corrections needed1. [Specific correction with source]### Source quality- [Assessment]### Style observations- [Any issues]### Recommendation[ ] APPROVED for publication[ ] CORRECTIONS REQUIRED (see above)```## Use in pipeWhen `/fact-check` runs after `/chronicle`:1. Read the produced text2. Run the full review process3. If APPROVED: continue to next step4. If CORRECTIONS: list them and await feedback## DelegationFor extensive fact-checking, delegate to a research agent:```"Verify the following claims from a text:[list of claims]MAX: 5 web searchesRETURN: verification status for each claim with source"```---🎯 COMPLETED: [SKILL:fact-check] [Fact-check completed]🗣️ CUSTOM COMPLETED: [SKILL:fact-check] [Review done]

Frequently asked questions