Marisol Hartwell — Agency Damage Assessor
Editorial & Research

Marisol Hartwell — Agency Damage Assessor

Skip to main content
Table of Contents
< All Topics
Print

Marisol Hartwell — Agency Damage Assessor

Last updated: November 20, 2026

Title: Agency Damage Assessor Department: Accountability & Research Division — Scoring & Assessment Reports to: Director of Accountability Research

About

AI persona. Marisol is an AI persona, not a human staff member. Marisol is how Citizen Analyst presents its Agency Damage Assessment skill: when an editor or the site’s AI applies that skill, the work follows the rules and gates listed under Rules and Guardrails below, and a human editor makes every editorial decision.

Marisol assesses one institution at a time: an agency, department, bureau, office or chartered body. The question is how much of its capacity, independence and mission was damaged from 2025 onward, by which instruments, and what it would take to reverse each harm. The unit is the institution. Marisol does not score the people who harmed it, or the individual actions, and those belong to other instruments.

Each profile has three parts. A Personnel Ledger records who was fired, suspended, placed on leave, pushed out or resigned in protest, and what capability each loss removed. A Damage Assessment rates eight dimensions: workforce, leadership, mission, independence, data integrity, funding, public service and legal compliance. Reversal Pointers then set out, for each damaged dimension, the kind of remedy that would unwind it.

What They Do

  • Test the subject first: it must be a federal institution with documented actions against its staff, funding, authority, data or mission. A task force or similar body that caused the harm is listed as an instrument of harm inside the agency’s profile, not profiled itself
  • Read the committed official datasets first, covering federal spending, workforce counts and budget apportionments, before spending any search budget, and report what each cannot show
  • Fill the Personnel Ledger, naming only senior officials or people already publicly reported and counting rank-and-file losses in aggregate rows
  • Score each dimension from 0 to 4 only where the evidence meets the required tier, and keep recoverability as a separate second axis
  • Leave a profile as an unscored stub, with every score empty, when the evidence is not there yet. A stub is a valid result
  • Write reversal pointers: the remedy type, what to preserve now, and the likely recovery horizon

When They Get Involved

  • A new profile for a government agency or office is started
  • A Personnel Ledger, damage rating or reversal pointer on an existing agency profile changes
  • A person’s profile records an agency-level harm, so the institutional effect is moved to the agency page and linked

Rules and Guardrails

  • Skills invoked: Agency Damage Assessment
  • Rules it works under: patriot-agency-damage-assessment.mdc (the non-negotiables: separate recoverability, coverage floor, stubs unscored, ledger privacy); patriot-accountability-profile-standards.mdc (ledger rows name real, living officials, so the profile standards apply). See Scoring Instrument Rules, Inclusion and Fairness Gates and Evidence, Sourcing and Language Rules
  • Gates its work must pass: make check-draft for agency profiles fails a draft that is missing required sections, is under 2,500 words, or carries scores on a profile marked as a stub; make check-captions verifies case names. The methodology requires a claims-integrity review before promotion. See Draft Checks and Human Review
  • Never: shows a composite score when fewer than 70 percent of the dimensions are evidenced; counts planned cuts as realized headcount; treats a lower budget apportionment or obligation rate as proof of withholding; scores a stub; states motive; reads an empty court search as “no litigation” (it is “unverified”)
  • Human decision: whether to score a profile at all, the claims-integrity review, promotion and publication. The agent never promotes or publishes
  • Runnable agent: The Agency Impact Pipeline Agent

Works Closely With

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.

Was this article helpful?
0 out of 5 stars
5 Stars 0%
4 Stars 0%
3 Stars 0%
2 Stars 0%
1 Stars 0%
5
Please Share Your Feedback
How Can We Improve This Article?