agentleFS
Sign inStart free
jezwebjezweb/vite-flare-starter/skills/fact-check/SKILL.md

fact-check

Skill49 starsChanged 3 months agoMITView on GitHub

Verify a specific claim by finding multiple authoritative sources that confirm or contradict it. Use when the user asks to fact-check, verify, or check whether something is true.

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 31 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:

    If the claim is partly true, be specific about which parts.

    line 41
  2. Distinctive:

    Don't trust a single source, even an authoritative one, for contested claims.

    line 44
  3. Distinctive:

    Don't conflate "common knowledge" with verified fact.

    line 45
  4. Distinctive:

    Don't fabricate confidence — uncertainty is a valid finding.

    line 46

“Common” is measured: kept by at least 6 in 10 of the 31 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

When to use

The user has stated a specific claim and wants to know if it's accurate. Examples:

  • "Is it true that X?"
  • "Fact-check this: ..."
  • "Did Y actually happen?"

Steps

  1. Identify the claim — extract the precise factual assertion. If the claim is fuzzy ("apparently X is good"), use ask_questions to nail down what specifically to verify.
  2. Search for primary sources — use web_search with the claim phrased neutrally. Look for:
    • Official sources (government sites, academic papers, the organisation involved)
    • Reputable journalism (major outlets with editorial standards)
    • Primary data (statistics agencies, public records)
  3. Avoid these as sole sources:
    • Social media posts (use only as leads to follow up)
    • Single-source blogs without citations
    • Anything that quotes the claim back without independent verification
  4. Cross-reference — find at least 2-3 independent sources. If they all trace back to one origin, you only have one source.
  5. Report verdict — use show_alert with the appropriate type:
    • success — Confirmed (multiple authoritative sources agree)
    • warning — Partly true / context needed
    • error — False (sources contradict the claim)
    • info — Unverifiable (insufficient evidence either way)
  6. Provide evidence — list the sources with what each one says. Use markdown links.

Style

  • Be neutral. Avoid editorialising.
  • Distinguish "X is false" from "I couldn't find evidence for X" — these are different.
  • If the claim is partly true, be specific about which parts.

What not to do

  • Don't trust a single source, even an authoritative one, for contested claims.
  • Don't conflate "common knowledge" with verified fact.
  • Don't fabricate confidence — uncertainty is a valid finding.

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.

Check your setup, free