Open-Source Verification Lab

A reproducible method for defining claims, resolving entities, preserving public material, corroborating findings, testing contrary evidence, and publishing qualified conclusions.

Public-interest problem

Search results collapse primary records, copied allegations, later corrections, and inference into one undifferentiated stream.

Primary users

Journalists, civic researchers, community historians, policy analysts, students, and document-led investigators.

Current tools & deliverables

What the project produces

Investigation brief
Source and entity map
Evidence log CSV
Confidence vocabulary
Correction and version record

Operating workflow

01Write a falsifiable question.
02Resolve people, entities, jurisdictions, dates, and identifiers.
03Map primary record systems before broad searching.
04Preserve provenance and distinguish leads from evidence.
05Corroborate independently and seek contrary records.
06Publish the support, limits, and correction route.

Data, privacy & limitations

Builders and working templates operate in the browser. CTI does not receive the information entered. Published work follows source provenance, PII minimization, lawful-access, right-of-reply, correction, and anti-doxxing standards.

Open sources can be incomplete, stale, decontextualized, or misidentified. CTI prohibits unauthorized access, pretexting, stalking, and doxxing.

Roadmap

Next development stage

Add verification exercises, source-quality rubrics, redaction demonstrations, and peer-review checklists.

Use the project

Move from the project file to the working method.

Open related guide