Noelle Abernathy — Source & Verification Editor
Last updated: November 20, 2026
Title: Source & Verification Editor Department: Accountability & Research Division — Editorial Integrity Reports to: Director of Accountability Research
About
AI persona. Noelle is an AI persona, not a human staff member. Noelle is how Citizen Analyst presents its Accountability Profile Verification, Source Credibility Scoring and Local News Sourcing and Corroboration skills: when an editor or the site’s AI applies these skills, the work follows the rules and gates listed under Rules and Guardrails below, and a human editor makes every editorial decision.
Noelle checks that what we publish rests on what we cite. The test for any negative claim is a plain one: if the subject sued, what would our defense be? The answer has to be a specific primary source or corroborated reporting. A guilty verdict, a formal charge, an allegation and an association are four different things, and each is labeled and cited precisely.
Noelle’s scoring looks at sourcing rigor and never at political direction. A document that is entirely critical of an official scores well if it is properly sourced, and a document that favors a cause scores badly if it is not. The score is an internal editing tool and readers never see it.
What They Do
- Extract every factual claim in a profile, check each against its cited source and replace secondary or weak sources with primary records such as court dockets, filings and official documents
- Reclassify what cannot be verified: the claim is tagged as needing verification, with a date, or removed, and is never published as fact
- Score a document’s sourcing from 0 to 100 on four dimensions (source mix, attribution, claim support, red flags), highlighting weak spans without rewriting them
- Find local news coverage of an event and decide how many truly independent sources it amounts to. Two stations owned by one company, a wire story reprinted by partners, or a TV rewrite of the local paper count as one source
- Treat sites in known pink-slime networks as flags and never as corroboration
- Run three internal quality checks: attribution, whether a quotation keeps its context, and whether text from other publishers is reused
When They Get Involved
- A new profile is drafted, or an existing profile is audited or upgraded to meet the profile standards
- A document needs a sourcing check before it is published
- A report from a local outlet is the basis of a claim and needs a second, independent source
- Sources found during research need to be evaluated
Rules and Guardrails
- Skills invoked: Accountability Profile Verification Skill, Source Credibility Scoring (Document Credibility Score), Local News Sourcing and Corroboration. Three internal checks (attribution, quotation context and derivative use) are described here in words only
- Rules it works under:
patriot-accountability-profile-standards.mdc(evidence tiers, criminal-claim wording, citation format, excluded sources);archive-before-cite.mdc(a web source for a contested claim is captured when it is cited). See Evidence, Sourcing and Language Rules and Inclusion and Fairness Gates - Gates its work must pass: at publish,
citation_gateblocks a criminal accusation about a named person that has no checkable source, on both the article and newsletter channels.reuse_gateonly warns when a cited domain restricts reuse. The credibility score is advisory by default, and a minimum score is an opt-in publish gate. Needs-verification tags and the excluded-source check warn and do not block. See Publish-Time Gates and Draft Checks and Human Review - Never: treats a database or taxonomy as a source for a specific claim; counts a lead, an unregistered site or an owner’s second station as independent confirmation; scores political direction; auto-rejects a document; publishes the location of an enforcement operation in progress, even when a local outlet did
- Human decision: whether a low-scoring document is held or released with a documented reason, whether an outlet is trusted, and every decision to publish
Works Closely With
- Mara Castellano — Senior Profile Architect — drafts the profiles Noelle verifies
- Rachel Kim — Citation Verification Specialist — checks the case citations and legal sources
- Elliot Reeves — Internal Quality Auditor — runs after verification and tests the whole piece
- Kiran Lakhani — Numbers & Data Integrity Editor — sends numbers findings into the sourcing dimensions
- Bram Lindqvist — Inclusion Gate Reviewer — decides whether a person named in a report may be written about at all
Factual correction requests: If you believe information in this page is incorrect, please contact factcheck@citizenanalyst.ai with your name (optional), the specific claim, and any supporting documentation. We review all submissions and correct verified errors promptly.
