agentleFS
PingPingEPingPingE/ai-content-agents/.claude/skills/fact-check/SKILL.md

fact-check

Skill2 starsChanged 9 months agoMITView on GitHub

Verify factual claims in content

39 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use deep-research, seo and 2 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • NVIDIASkill

    deep-research

    Queue deep, multi-step research and analysis tasks to the DeepAgents worker. Supports execution depth presets, request-specific rubrics, SubAgent delegation, and RubricMiddleware cross-validation.

    Also in 4 projects that use this

  • nilssonrSkill

    seo

    Full SEO workflow for auditing pages and creating SEO-optimized content. Provides checklists with educational explanations aligned with Google Search Central documentation.

    Also in 3 projects that use this

  • NVIDIASkill

    research

    Before designing a feature, research how comparable libraries solve the problem, check best practices, known pitfalls, and upstream changes. Use when starting a new feature, investigating an unfamiliar area, or the user asks to research something.

    Also in 3 projects that use this

  • BlockSkill

    frontend-design

    Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 32 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Verify factual claims in:

    line 16
  2. Distinctive:

    Read the content and identify every factual statement.

    line 32
  3. Distinctive:

    Use WebSearch to find:

    line 46
  4. Distinctive:

    Avoid: blog posts, social media, unattributed aggregators

    line 51

“Common” is measured: kept by at least 6 in 10 of the 32 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Fact Check

Verify factual claims in:

Target: $ARGUMENTS

Check every:

  1. Statistic and number
  2. Factual claim
  3. Attribution and quote
  4. Technical description
  5. Referenced link

Report as: verified / unverified / incorrect with sources.

Verification Pipeline

Step 1: Extract Claims

Read the content and identify every factual statement. Categorize as:

  • Statistics — percentages, numbers, survey results
  • Historical facts — dates, events, names
  • Quotes — attributed statements
  • Technical claims — API behavior, code functionality, software capabilities
  • Predictions — future-oriented statements

Step 2: Classify Each Claim

For each claim, determine:

  • Claim type — which category above?
  • Verifiability — can this be fact-checked?
  • Priority — how important is accuracy for this claim?

Step 3: Verify Each Claim

Use WebSearch to find:

  1. Primary sources — original research, official docs, direct data
  2. Secondary sources — peer-reviewed articles, reputable news
  3. Tertiary sources — Wikipedia (starting point only, not final source)

Avoid: blog posts, social media, unattributed aggregators

Step 4: Report Findings

For each claim, document:

  • Location in content
  • Status (Verified/Likely/Unverified/Incorrect)
  • Source URL
  • Notes on any issues

Source Quality Tiers

Tier 1 (Best): Official documentation, peer-reviewed research, government data, primary sources Tier 2 (Good): Reputable news outlets, industry reports, verified expert statements Tier 3 (Starting point): Wikipedia, aggregators (verify with Tier 1/2 sources) Tier 4 (Avoid): Blogs, social media, opinion pieces, uncited claims

Confidence Levels

  • Verified — found authoritative source, claim is accurate
  • Likely — multiple indirect sources confirm, but no primary source found
  • Unverified — cannot find source to confirm or deny
  • Incorrect — found authoritative source that contradicts claim

Report Format

ClaimLocationTypeStatusSourceNotes
"75% of developers use Git"Para 2StatisticVerifiedStack Overflow Developer Survey 2025Current and accurate
"Einstein said innovation requires..."IntroQuoteIncorrectQuote InvestigatorMisattribution — actually Mark Twain
"Next.js 15 uses React 19"Para 5TechnicalVerifiedNext.js official docsAccurate as of Feb 2026
"80% of startups fail in year 1"Para 3StatisticUnverifiedNo primary source foundCommon claim but unsubstantiated

Summary Statistics

  • Total claims checked: X
  • Verified: Y (Z%)
  • Likely: Y (Z%)
  • Unverified: Y (Z%)
  • Incorrect: Y (Z%)

Action Required

Critical (must fix):

  1. [Incorrect claim] — replace with: [correct information] from [source]
  2. [Another incorrect claim] — replace with: [correct information]

Recommended (add sources):

  1. [Unverified claim] — add citation to: [suggested source]
  2. [Another unverified claim] — add citation or remove claim

Low priority:

  1. [Outdated but not wrong] — consider updating to more recent data

Quality Checks

Before submitting report:

  • Checked every statistic and number
  • Verified all quotes and attributions
  • Tested all URLs (no 404s)
  • Flagged outdated information (2+ years old)
  • Provided source URLs for all verified claims
  • Offered correct information for all incorrect claims

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.