Last updated:

AI Dorks — Claim Verification

Claude prompts for checking claims and content — fetching primary sources for a sourced verdict, adjudicating evidence you provide, and extracting clean citable text.

Public-source, authorized use. These are prompt-engineering aids, not jailbreaks. Use them on subjects and infrastructure you’re authorized to investigate; they keep to public sources, respect site terms, and exclude breach data and private-individual targeting. Paste a prompt into Claude, fill the highlighted fields, and have it show its work and cite sources.

Fact-check against primary sources with Claude

These prompts have Claude fetch the primary and official pages that bear on a claim, grade each source, and conclude Supported, Mixed, Refuted, or Unverifiable with cited URLs — or adjudicate a set of sources you paste without inventing new facts. A trafilatura prompt pulls clean, citable article text from a page. Claude can also generate Google dork and reverse-image queries to locate the primary sources it then grades.

Frequently asked questions

How is this different from just asking Claude?

These prompts force source discipline: fetch primary pages, cite every claim, grade source quality, and reach an explicit verdict — rather than answering from memory.

Can it adjudicate without internet?

Yes — one prompt weighs only the sources you paste, flags fallacies, and gives a confidence, adding no outside facts.

Does it respect site terms?

Yes — fetching is for public content with robots.txt and terms respected; it won’t bypass paywalls or logins.

How do I verify whether a photo is real?

Combine signals: check EXIF and any C2PA content credentials, reverse-image-search for earlier appearances, and test that shadows and weather match the claimed place and time. These prompts have Claude gather these and reach a graded, cited verdict.

What are good fact-checking tools?

Beyond AI prompts, fact-checkers rely on reverse image search, EXIF and metadata viewers, and primary-source databases. These prompts orchestrate the source-gathering and force Claude to cite every claim and grade each source.

Can AI dorking replace Google dorking for fact-checking?

They complement each other: Google operators find candidate primary sources, then these prompts have Claude fetch, grade and cite them for a verdict and reverse-image-check the visuals. See the Google Dorks list for the operators.