For research teams
A branching mistake can silently remove a variable from every record collected after an amendment. FormPath checks the released definition, compares analyzed behavior across revisions, and produces evidence you can retain beside change documentation.
Bring a SurveyJS definition, REDCap data dictionary, or FHIR Questionnaire—not participant responses. The report separates defects it can demonstrate from logic that needs the running application or a manual check.
Study lifecycle
Check that branching values still exist, supported fields can appear, and supported conditions were seen both true and false where possible. The report separates these answer checks from full respondent journeys and does not audit every required-field behavior.
Separate newly proven regressions from intended field and rule changes. Preserve the report beside amendment documentation.
Compare site dictionaries after local adaptations so answer recoding does not leave an old branching value behind.
These files were collected to exercise varied definitions and adapters, not to estimate a population defect rate. The public ledger preserves every denominator and evidence boundary.
Inspect the retained reviewBefore you begin
Paste or select a SurveyJS JSON definition, a REDCap data dictionary CSV, or a FHIR Questionnaire JSON file. FormPath reads the definition itself; it does not need a live project connection.
No. Analysis runs in a worker on this device. FormPath does not upload the definition or fetch remote choice lists. A report leaves the session only if you choose to copy or download it.
No. Do not include responses, PHI, credentials, API tokens, or secrets. A form definition is the intended input.
It means FormPath did not prove a defect within the supported, completed scope shown in that report. It does not certify the whole instrument. Unsupported rules, application-controlled behavior, and unfinished checks remain listed for review.
What leaves the browser
FormPath does not need study records, participant responses, or direct REDCap access. Export only the report or generated test answers you choose to keep.
Working SurveyJS answer combinations are replayed in SurveyJS. Supported impossible rules use finite proofs whose encoding is sample-checked against SurveyJS for that definition. REDCap and FHIR rules are translated into a tested internal format and are labeled as translated in the report.