geo-fact-checker_skill

This skill verifies factual claims for content, cites reliable sources, and updates data with evidence to boost AI trust and citation readiness.
  • Python

2.6k

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

3 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill geo-fact-checker

  • _meta.json287 B
  • SKILL.md21.4 KB

Overview

This skill is a GEO-focused fact-checking and evidence-collection assistant designed to improve the factual reliability and citation readiness of written content. It targets numbers, dates, rankings, market-share claims, competitor data, quotes, and other high-impact statements to make content safe to cite. The skill prioritizes accuracy, transparency, and traceability over stylistic polish.

How this skill works

I scan the provided text to extract and classify factual claims (numeric statistics, dates, rankings, competitor mentions, quotes). For each claim I plan an evidence search, query authoritative sources, and compare findings to the original text. I then label each claim (verified, partially_verified, outdated, contradicted, uncertain) and propose corrected wording with clear citation-ready summaries of the supporting sources.

When to use it

  • When content contains numbers, dates, percentages, rankings, or market-share statements.
  • When preparing reports, comparison pages, landing pages, or data-driven articles that need reliable citations.
  • When validating competitor statements, product claims, or third-party quotes before publishing.
  • When updating older content to reflect current data or when AI outputs require traceable evidence.
  • When the audience or decision is high-stakes (B2B procurement, financial, legal, health-adjacent).

Best practices

  • Specify a time horizon (e.g., ‘as of 2026’) and region to narrow searches and avoid ambiguous claims.
  • Prioritize authoritative sources (official reports, government data, recognized research bodies) and cite domains and publication years.
  • Document assumptions and tolerances (units, rounding thresholds, scope) so reviewers can audit decisions.
  • Flag uncertainty explicitly rather than over-asserting: use cautious language when evidence is mixed or absent.
  • Provide both the status label and a concise recommended rewrite for any non-verified claim.

Example use cases

  • Audit a landing page that claims ‘#1 in market share’ before publishing to avoid legal exposure.
  • Verify user counts, revenue figures, or growth rates in an investor-facing report with dated sources.
  • Check and update a ‘Top 10 tools’ comparison page to ensure rankings and features match current data.
  • Validate quoted statistics from press releases and replace unsupported numbers with sourced alternatives.
  • Convert draft blog posts into citation-ready articles suitable for AI-driven search and GEO.

FAQ

Prefer official company reports, government agencies, peer-reviewed research, and recognized industry analysts; avoid single, low-credibility blogs when verifying high-impact claims.

How do you handle conflicting sources?

I present the conflict, assess source credibility and recency, label the claim as partially_verified or uncertain, and recommend cautious phrasing or removal if needed.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational
geo-fact-checker skill by openclaw/skills | VeilStrat