Verify a factual claim with independent evidence

Why use this

1. A reader checking a factual assertion may find many apparently independent articles that all repeat one original source. Later corrections or narrower definitions can also change whether the precise claim remains supported.

2. A source-origin review could separate corroboration from repetition and show where evidence contradicts or qualifies the assertion. The reader would receive a bounded assessment rather than a confidence label based merely on the number of links.

Why an agent / simpler alternative

Useful judgment

Compare exact claim scope, trace shared origins and weigh relevant contrary evidence.

Simpler alternative

Check the primary source directly when the claim is narrow and its evidence is straightforward.

Unsuitable when

The question is a real-time high-stakes determination requiring specialist or immediate authoritative judgment.

Proposed inputs

1. Name

   Exact assertion

   Owner

   Requester

   Minimum scope

   One claim with dates, entities and important qualifiers

   Quality check

   Avoid changing the claim into an easier version to verify.

2. Name

   Evidence packet and search scope

   Owner

   Research owner

   Minimum scope

   Permitted sources, dated passages and known corrections

   Quality check

   Preserve original provenance and distinguish primary evidence from commentary.

Access and setup — proposed or unknown

Proposed minimum

1. Read-only supplied evidence and an offline claim-evidence matrix.

Compatibility unknowns

1. Unknown: exact runtime, search coverage, source access and revision-history availability.

External actions requiring approval

1. Contacting sources, publishing consequential allegations or sharing restricted evidence.

Expected deliverable and review checks

Proposed artifact

Cited assessment with supporting/contrary evidence, origin groups and explicit unresolved questions.

Critical checks

1. Evidence addresses the exact assertion and time scope.

2. Shared-source articles are not treated as independent confirmation.

Missing output evidence

No new workflow run or output inspection is performed for this editorial layer. The linked implementation evidence retains the original source-specific outcome and artifact limits.

Failure signals and proposed responses

1. Signal

   Several reports trace back to one unverified announcement.

   Why it matters

   Counting them separately would inflate confidence.

   Proposed response

   Group their shared origin and identify what independent evidence is missing.

2. Signal

   A later correction changes a key number or qualifier.

   Why it matters

   An older quote may no longer support the claim.

   Proposed response

   Include the correction and revise the assessment’s scope.

Effort and cost — unknown

Setup work

1. Define the assertion and source-independence criteria.

Ongoing review

1. Check revisions and keep confidence tied to available evidence.

Unknowns

1. Unknown: source completeness, access costs and unresolved evidentiary quality.

First check — proposed, unexecuted

Status

proposed_unexecuted

Synthetic fixture

Invented reports repeat one announcement, while a later primary-source correction narrows the original statement.

Acceptance checks

1. The repeats remain one evidence origin.

2. The correction changes the proposed assessment.

Stop conditions

1. A missing primary source or high-stakes conclusion must be guessed.

Limits

1. Editorial verification structure; source quantity alone does not establish truth or independent corroboration.