App access check complete. Continue to the page below.
An AI-assisted political research tool that checks candidate identity, records source links and automated review notes, and applies Cincinnatus's disclosed scoring rubric.
Identity first
State, office, district, election year, and current web sources are used to avoid analyzing the wrong person.
Separate AI check
A separate verification pass must approve identity and core facts before a new score is published.
Traceable evidence
High-risk claims such as campaign-finance concerns, votes, associations, and stock trades require direct source links.
Cincinnatus is named after the Roman statesman associated with the citizen-servant ideal. The app is designed to make political research easier to inspect by bringing identity, public-record research, citations, and a consistent scoring framework into one report.
The +100 to -100 score is a values-based rubric, not an objective or politically neutral measurement. It reflects the ten criteria shown in the app, including moral character, family stability, fiscal responsibility, practical experience, national security, and other stated values. Users can inspect the evidence and individual trait explanations rather than relying on the total score alone.
The Career Politician trait assigns -10 for 8 or more cumulative years in elected political office, regardless of private-sector experience or other qualities. The existing scope counts elected service, not appointed service. Service duration must be supported by sources; unknown or unverified duration stays unknown and this trait is held at 0. Below eight years, the other disclosed criteria for this trait remain in effect. Previously saved reports keep their recorded explanations until refreshed. Unsupported or inconsistent aggregate scores are withheld in the displayed report without rewriting the saved evidence or replacing an unknown score with zero.
Data quality is treated separately from the rubric. New candidate searches first resolve the exact identity. When more than one person is plausible, Cincinnatus asks the user to choose instead of silently guessing. A separate AI pass then reviews the draft report; this is not independent fact-checking. Low-confidence or identity-mismatched reports are blocked from publication.
Supplemental sections can include campaign-finance research, recorded votes, recent quotes, a dark-money/conflict-risk section, a foreign-intervention index, stock-trading conflict research, and a narrowly sourced personal-association check. Unsupported high-risk items are omitted rather than presented as established fact.
The ballot tool sends the supplied address to an AI research service and saves it to your account. It returns research links from a single pass. Address eligibility, filing and ballot qualification are not independently verified, so these results are labeled as needing confirmation.
Cincinnatus is an independent project and is not affiliated with a political party, PAC, or campaign. AI-assisted research can still contain errors, so the app exposes sources and recorded AI review notes so users can check important claims themselves. An automated review is not an independent fact-check.
A link is a starting point for review, not proof by itself. A separate AI pass checks candidate reports; it can still make mistakes. When that pass rejects evidence in a high-risk section, the section's earlier summary and score are withheld. Missing age or service history stays unknown.
Cached reports can be reused for up to two days for election races or seven days for officeholder research. Automated refresh is limited to one eligible report per day and does not guarantee that every stored report is current. Check the report date and original records.
Report a correction with the affected report and a supporting source.