Structured, evidence-tiered classification that evaluates where a government spokesperson’s in-role communications fall on the spectrum from legitimate advocacy and message discipline to knowing misinformation designed to sow public confusion and distrust. Produces a Spokesperson Misinformation Assessment (SMA) with a transparent advocacy-vs-misinformation gate, a 1-5 Misinformation Severity Score (MSS), an ordinal comparative baseline against a bipartisan reference class of press secretaries, and layered defamation-law protections. Use when creating or updating a spokesperson/communications-official profile that already carries per-statement veracity ratings. Composes with spokesperson-profile-builder (runs before — writes statements + individual veracity ratings), accountability-profile-verification (runs before), and patriot-sanity-check (runs after).

This skill is the communications-side sibling of malice-evaluator. Where the Democratic Malice Assessment (DMA) scores documented actions for democratic-mechanism subversion, the SMA scores documented statements for the transition from advocacy to misinformation. It reuses DMA’s gate-then-score architecture and defamation framework, and reuses the deviation-from-baseline pattern of international-mining-deal-specialist for its comparative layer.

---

When to Use This Skill

Run the SMA when ANY of the following apply:

1. Creating or updating a profile with profile_type: spokesperson (or subject_comms_role set), after spokesperson-profile-builder has written the Significant Public Statements and Overall Veracity Track Record sections
2. A spokesperson profile documents 2+ in-role statements with institutional fact-check ratings
3. Processing any spokesperson profile with category: trump-administration (each requires either an SMA designation OR an explicit “No SMA” transparency block)
4. Reviewing spokesperson profiles during periodic audits for SMA coverage

Do NOT run this skill for:

• Profiles still in _drafts/ status
• Profiles that lack the spokesperson template (no Significant Public Statements / Overall Veracity Track Record sections) — the SMA consumes those per-statement ratings as input and cannot run without them. A subject who was a spokesperson but whose profile is accountability-typed needs a spokesperson (-comms) profile built first.
• Profiles without institutional fact-check citations for any statement (SMA is not eligible; see Section 3 evidence gating)

---

Constitutional and Philosophical Foundation

The SMA framework draws from three domains:

Legal concepts:

• “Actual malice” in New York Times v. Sullivan (1964) — knowing falsity or reckless disregard for truth. Government spokespeople are quintessential public officials.
• The distinction between factual claims and evaluative judgments (Milkovich v. Lorain Journal, 1990) — an MSS is an evaluation of communicative intent, cited separately from the underlying fact-check.
• Government-speech doctrine: an official speaking in their communications role speaks for the government, not as a private citizen — the basis for the Tenure-Window Guard.

Disinformation research:

• Glenn Kessler / Washington Post Fact Checker Pinocchio scale — graduated severity from selective-truth shading (1 Pinocchio) to “whoppers” (4), and the Bottomless Pinocchio (a false claim repeated 20+ times after being debunked) — the precedent for treating repetition-after-correction as an aggravating factor.
• RAND “Firehose of Falsehood” model — high-volume, multi-channel, rapid, repetitive messaging with no commitment to objective reality or internal consistency; the goal is to entertain, confuse, and overwhelm rather than persuade.
• The illusory truth effect and “truth apathy” / learned helplessness — repeated exposure to falsehoods erodes the public’s baseline trust in verification itself. This is why institutional-trust targeting and verification delegitimization are weighted more heavily than isolated factual errors: they attack the mechanisms of shared reality, not just individual facts.

Civic principle:

• Message discipline in service of an elected President’s agenda is a legitimate, expected function of every administration’s communications staff, of both parties. The SMA exists to mark the boundary where that function is abandoned in favor of knowingly corrosive falsehood — not to penalize advocacy.

---

Section 1: The Advocacy vs. Misinformation Gate

Purpose

This gate runs FIRST, before any scoring. It separates legitimate advocacy and message discipline (protected, expected, even when one-sided or aggressive) from misinformation (statements that require the speaker to know the claim is false or misleading and whose function is to create confusion or distrust). Every statement assessed passes through this gate with a transparent, reader-facing classification.

The Tenure-Window Eligibility Guard (categorical exclusion — runs before the gate)

Only statements made while the person held the spokesperson/communications role and was speaking in that official capacity are eligible for SMA scoring or baseline inclusion. Statements made before they assumed the role and after they left it are categorically excluded — almost every former press secretary becomes a paid television commentator (Fox, CNN, MSNBC, etc.), and that speech is private-figure opinion, not government messaging. Scoring it would be methodologically wrong (it is not the administration speaking) and would forfeit the NYT v. Sullivan public-official framing the defamation layer depends on.

Implementation:

• Every profile and every baseline row carries explicit subject_tenure_start / subject_tenure_end (hardcoded ISO dates; subject_tenure_end: ongoing for incumbents).
• Date-filter every candidate statement to [subject_tenure_start, subject_tenure_end] before applying the gate. Out-of-window statements are dropped with a one-line note (“Excluded: post-tenure commentary, [date]”) and never scored.
• Fact-checker scorecards are lifetime, not tenure-windowed. A raw PolitiFact “report card” spans a person’s whole public career; it cannot be used wholesale for anyone with a pre/post-government media career. Filter individual ratings by date to the tenure window.

The Boundary Test

For every in-window statement under assessment, answer:

“Could this statement serve legitimate message discipline or favorable spin — even if one-sided — or does it require the speaker to know the claim is false or misleading?”

Classification	Definition	Example
Advocacy	Favorable framing, emphasis, or spin that does not require asserting a known falsehood	“This is the most pro-worker agenda in a generation” — contestable opinion, not a false fact
Misinformation	Assertion of a factually false or materially misleading claim, in circumstances indicating the speaker knew or should have known, whose function is to confuse or erode trust	Repeating a specific statistic after the agency’s own data publicly refuted it
Borderline	Could be either, depending on documented knowledge and pattern — requires the five-factor test below	A blame attribution that omits a material fact the speaker had access to

The Five Distinguishing Factors

When a statement falls in the borderline zone, apply these factors. Any 3 of 5 tips the classification to misinformation. These map directly to the five misinformation categories in Section 2.

1. Debunked repetition — the claim was repeated after a formal correction or institutional fact-check (Bottomless-Pinocchio pattern).
2. Firehose pattern — the same false framing was pushed across multiple channels or coordinated across multiple spokespeople simultaneously.
3. Internal contradiction — the statement conflicts with the speaker’s own prior on-record statements or with the agency’s own published data.
4. Institutional-trust targeting — the claim’s function is to delegitimize courts, election systems, the press, or agency data, rather than to persuade on policy merits.
5. Verification delegitimization — the statement attacks the fact-checking or verification process itself (“fake news,” discrediting named outlets) rather than rebutting the substance.

Advocacy Safe Harbor (NEVER scored as misinformation)

The following are always classified as advocacy regardless of how one-sided:

• Favorable framing or characterization of true facts
• Declining to comment, or refusing to confirm/deny
• Disputing an opponent’s characterization or predicting a policy’s benefits
• An isolated factual error, corrected or not repeated (Kessler’s 1-Pinocchio zone)
• Rhetorical emphasis, hyperbole, or selective framing without factual falsity
• Aggressive or combative tone toward the press, absent a false factual claim
• Defending the President’s stated position or agenda as a matter of opinion
• Accurately relaying an administration position the spokesperson did not originate

Reader-Facing Transparency Output

Every SMA-scored statement MUST include a visible disclosure within its “Scored Statement” sub-section (see Section 8, Format A). The Advocacy vs. Misinformation Determination field provides this inline:

**Advocacy vs. Misinformation Determination:** **Misinformation.** [1-3 sentence
explanation citing the specific distinguishing factors met. End with: "The legitimate-
advocacy version of this message would be: [1 sentence describing the true-fact framing
available to the speaker]."]

For profiles receiving no SMA designation, see Section 8, Format B.

---

Section 2: The Five Misinformation Categories

Each scored statement is tagged with one or more categories. Categories 4 and 5 carry the heaviest weight because disinformation research shows they erode the public’s baseline trust in shared reality, not merely a single fact.

Category	Definition	Why it matters
Debunked Repetition	Repeating a claim after a formal correction or institutional fact-check	Repetition drives the illusory-truth effect; the Bottomless Pinocchio
Firehose Amplification	Same false framing across multiple channels / coordinated spokespeople	Volume and multi-channel repetition overwhelm correction (RAND)
Internal Contradiction	Claim conflicts with the speaker’s own record or the agency’s own data	Demonstrates knowledge of falsity; hardest to defend as good-faith error
Institutional-Trust Targeting	Claim’s function is to delegitimize courts, elections, press, or official data	Corrodes the mechanisms of shared reality — highest-weight harm
Verification Delegitimization	Attacks the fact-checking process itself rather than the substance	Inoculates the audience against all future correction — highest-weight harm

Category Definitions — Detailed

Debunked Repetition: The spokesperson restated a specific factual claim after an institutional fact-checker (PolitiFact, FactCheck.org, Snopes, AP, Reuters, WaPo Fact Checker) had rated it false and the rating was publicly available. The aggravating element is repetition after correction, not the original statement.

Firehose Amplification: The same false framing appears across the spokesperson’s briefings, social accounts, and TV hits within a compressed window, and/or is echoed simultaneously by other administration spokespeople. The pattern — not any single instance — is the harm.

Internal Contradiction: The statement conflicts with (a) the speaker’s own prior on-record statement, or (b) the publishing agency’s own data or testimony. This is the strongest available proxy for knowledge of falsity, since the contradicting fact was within the speaker’s own institution.

Institutional-Trust Targeting: The claim’s operative function is to make the audience distrust a verification institution — courts (“rogue judges”), election administration (“rigged”), the press (“enemy of the people”), or the agency’s own data — rather than to win a policy argument. Scored higher because the harm is to shared reality itself.

Verification Delegitimization: The statement attacks the fact-checking or journalistic verification process by name (“fake news,” “discredited outlet,” “the fact-checkers are partisan”) in lieu of rebutting the substance of a correction. Scored higher because it is designed to immunize the audience against all future corrections.

---

Section 3: Misinformation Severity Scoring Protocol

Each statement that passes the Advocacy vs. Misinformation Gate (classified as “Misinformation”) receives a Misinformation Severity Score (MSS) on a 1-5 scale.

Score	Label	Criteria	Knowledge Evidence Required
1	Reckless Mischaracterization	Factually false or misleading, no evidence the speaker knew	Institutional fact-check rating; no evidence of prior knowledge
2	Willful Inaccuracy	Speaker ignored an available correction or the agency’s own data before restating	Fact-check + documented availability of the correcting information
3	Knowing Distortion	Contemporaneous evidence the speaker knew the claim was false (contradicts own agency’s data/testimony)	Fact-check + own-institution contradiction or on-record prior statement
4	Directed Falsehood Campaign	Repeated the same debunked claim across venues after a formal correction	Fact-check + documented repetition-after-correction across 2+ venues
5	Systemic Disinformation Operation	Sustained, multi-topic, multi-channel campaign explicitly targeting institutional trust	Fact-check + sustained multi-topic pattern + institutional-trust targeting

Evidence-Tier Gating

The SMA requires an institutional fact-check citation (PolitiFact, FactCheck.org, Snopes, AP Fact Check, Reuters Fact Check, or WaPo Fact Checker) as the factual predicate for every scored statement — the same standard spokesperson-profile-builder already mandates. Map the evidence tier to the eligible score band, mirroring DMA:

• Fact-check documented (institutional rating exists): eligible for MSS 1-5
• Credibly reported but not formally fact-checked: caps at MSS 2 (note the pattern; do not score higher)
• Outcome/impression inference alone (confusion demonstrably resulted, but no fact-check and no knowledge evidence): caps at MSS 1 and cannot by itself establish the “knowing” element for MSS 3+
• No institutional fact-check citation: not SMA-eligible — do not score

Scoring Procedure

For each documented, in-window statement:

1. Apply the Tenure-Window Guard — if out-of-window, drop (not scored).
2. Apply the Advocacy vs. Misinformation Gate — if Advocacy, stop (no score; note in “What Is NOT Scored”).
3. Confirm the evidence tier — is there an institutional fact-check citation? If not, not eligible.
4. Tag the misinformation category/categories (Section 2).
5. Determine the knowledge level — using the Knowledge Evidence Standards (Section 4).
6. Assign the MSS — the score matching the demonstrated knowledge level, capped by the evidence tier.
7. Write the transparency block — explain why this is misinformation, not advocacy, and state the legitimate-advocacy alternative.

---

Section 4: Knowledge Evidence Standards

Knowledge — that the speaker knew or should have known the claim was false — is the hardest element to establish and the most important for legal protection. The SMA uses a graduated evidence hierarchy.

Evidence Patterns by MSS Level

MSS 1 (Reckless Mischaracterization) — fact-check only:

• An institutional fact-checker rated the statement false or misleading.
• No evidence the speaker was aware of the correcting information at the time.
• Example evidence: a single PolitiFact “False” rating with no prior public correction.

MSS 2 (Willful Inaccuracy) — ignored available correction:

• The correcting information (agency data, a prior fact-check, expert consensus) was publicly available before the statement.
• The speaker was in a position to be aware of it.
• Example evidence: a CBO score, agency press release, or prior fact-check published before the restatement.

MSS 3 (Knowing Distortion) — knowledge documented:

• The statement contradicts the speaker’s own prior on-record statement, or the publishing agency’s own data or testimony.
• Contemporaneous evidence shows awareness.
• Example evidence: the agency’s own published dataset, the speaker’s own earlier briefing transcript, sworn testimony.

MSS 4 (Directed Falsehood Campaign) — repetition after correction documented:

• The same debunked claim was repeated across 2+ venues (briefing, social, TV) after a formal correction was public.
• Example evidence: dated transcripts/posts showing the claim restated after the fact-check publication date (Bottomless-Pinocchio pattern).

MSS 5 (Systemic Disinformation Operation) — sustained campaign documented:

• Multiple MSS 3-4 statements across different topics, plus explicit institutional-trust targeting.
• Pattern sustained over an extended period, not a single episode.
• Example evidence: a documented multi-topic pattern with dated fact-checks plus statements delegitimizing courts/elections/press/data.

Knowledge Assessment Guardrails

• A false rating alone never establishes knowledge above MSS 1 — an institutional “False” rating shows the claim was false, not that the speaker knew.
• Advocacy motivation is not misinformation motivation — wanting the President’s agenda to prevail is legitimate; knowingly asserting falsehoods to confuse the public is not.
• Relaying vs. originating — a spokesperson relaying an administration position they did not originate is weighted lower than one who originated or personally elaborated the false claim.
• Statements must be contemporaneous — later admissions are relevant but weighted less than real-time evidence of knowledge.
• The “should have known” standard is anchored to the role — a press secretary is charged with knowledge of their own agency’s published data.

---

Section 5: Actor-Level Cumulative Designation

When a spokesperson accumulates multiple qualifying statements (scored MSS 1+), they earn a cumulative designation.

Designation	Threshold	Meaning
No SMA designation	Fewer than 3 statements at MSS 2+, and none at MSS 4+	Statements may include errors but do not demonstrate a knowing pattern
Pattern of Strategic Misinformation	3+ statements at MSS 2+ OR 1 statement at MSS 4+	Documented pattern of willful or knowing misinformation
Sustained Disinformation Campaign	5+ statements at MSS 3+ OR 2+ at MSS 5 OR a documented Bottomless-Pinocchio repeated-after-correction pattern	Sustained, multi-topic, multi-channel campaign eroding institutional trust

Cumulative Designation Rules

• The designation is placed in the profile’s SMA block (see Section 8).
• It must cite the specific statements and their scores that justify the designation.
• It must be updated when new evidence emerges (up or down).
• The comparative baseline (Section 6) contextualizes but does not override the designation.
• patriot-sanity-check validates that designations are proportionate.

---

Section 6: Comparative Veracity Baseline (Deviation Layer)

The SMA scores two distinct things; keep them separate:

• Veracity rate — how often this person was fact-checked false while in role. This is inherently comparative, and institutional fact-checkers already measure it, so it is baselined against a reference class.
• Misinformation intent — the Gate + categories + MSS above. This is individual and evidence-specific; it is NOT baselined — you cannot baseline “did they knowingly lie” against a distribution. The baseline only supplies context for the cumulative designation.

This layer reuses the deviation-from-baseline pattern of international-mining-deal-specialist: establish a normative baseline, then score the subject as an ordinal deviation band — never a false-precision percentile or standard deviation.

Mechanics

1. Reference class = press secretaries and senior communications officials across administrations, explicitly bipartisan (e.g., Fleischer, Perino [Bush]; Gibbs, Carney, Earnest [Obama]; Spicer, Sanders, McEnany, Grisham [Trump 1.0]; Psaki, Jean-Pierre [Biden]; the current Trump 2.0 corps). The bipartisan class is itself the anti-bias and defamation defense — a subject is measured against both parties’ press secretaries. The reference data lives in spokesperson-conduct-baseline.md.
2. Segment by role tier — White House press secretary vs. agency spokesperson — to match SMA severity tiers. Do not over-segment (small n).
3. Score the subject as an ordinal band vs. the tier norm:

• within-norm — fact-check volume and false-rating share comparable to the reference class
• elevated — visibly above the reference class on false-rating share or repetition-after-correction
• outlier — far above any reference-class member on the tier

1. Selection-bias disclosure (mandatory). Institutional fact-checkers check the controversial claims, not a random sample, so this is an ordinal reference band, not a rate. Cite raw counts (e.g., “N institutional false ratings in the first M months in role”), never a computed “lie rate.” Carries the baseline guardrail verbatim: deviation is a signal for scrutiny, not proof.

The band feeds the “Comparative context” bullet in the profile’s Overall Veracity Track Record section and appears as the “Comparative Baseline” line in the SMA block. It contextualizes — it does not raise or lower — the MSS-driven cumulative designation.

---

Section 7: Defamation Law Protection Framework

The SMA maximizes legal protection against civil litigation by aligning every structural choice with established US defamation law. It reuses the DMA framework.

Layered Legal Defenses

Defense	How the SMA Implements It
Truth (Substantial Truth)	Every scored statement requires an institutional fact-check citation as its factual predicate. The predicate — “on [date], [outlet] rated this claim false” — is literally true. Truth is an absolute defense.
Fair Report Privilege	The SMA reports on public briefings, on-record statements, and published institutional fact-checks. Accurate reporting of public statements and public fact-checks is privileged.
Qualified Privilege	Accountability journalism and civic education serve a legitimate public interest, defeated only by actual malice, which the evidentiary standards prevent.
Opinion Privilege	The MSS and the “knowing” determination are evaluative judgments about communicative intent — closer to opinion than fact under Milkovich. The factual predicate (the fact-check) is cited separately.
Public Official Doctrine	Government spokespeople are public officials who must prove actual malice (NYT v. Sullivan) to prevail. The Tenure-Window Guard preserves this by excluding their private-citizen (post-tenure) speech.

The Tenure-Window Guard as a Legal Safeguard

Excluding pre- and post-tenure speech is not only methodological hygiene — it is a legal safeguard. A former press secretary’s Fox/CNN commentary is the speech of a private commentator, potentially a private figure for defamation purposes. Scoring it would both mischaracterize it as government messaging and weaken the Sullivan public-official framing. Only in-capacity, in-window government speech is scored.

The “Misinformation” Terminology — Legal Nuance

The SMA assesses the statement’s function and the speaker’s knowledge, not the platform’s animus:

• Platform’s state of mind: a good-faith analytical framework applied uniformly to both parties’ spokespeople. No personal animus. Defeats “actual malice” claims against the platform.
• Subject’s state of mind: assessing that a statement was knowing misinformation is an evaluation of communicative intent, which courts treat as closer to protected opinion than actionable fact.
• Factual predicates separated: the SMA separates the factual predicate (the cited institutional fact-check) from the evaluative conclusion (the MSS). The predicate must be true; the conclusion is protected opinion.

Design Choices Reducing Litigation Exposure

1. “Assessment” and “determination” language — not “accusation” of lying.
2. Published Advocacy Safe Harbor — demonstrates good faith and that ordinary spin is protected.
3. Bipartisan reference class — demonstrates lack of partisan animus.
4. Transparent methodology — exact criteria published; reviewable and rebuttable.
5. Evidence floor — MSS 3+ requires an institutional fact-check plus corroborating knowledge evidence.
6. Tenure-Window Guard — only in-role government speech is scored.
7. Proportionality controls — Gate + patriot-sanity-check prevent overreach.

Required Legal Disclaimer

Every SMA-bearing profile includes a “Legal Disclaimer” sub-section at the end of the SMA block (see Section 8, Format A):

### Legal Disclaimer

The Spokesperson Misinformation Assessment is an analytical framework applying defined
criteria to documented, in-role public statements by a government spokesperson. Each score
rests on a cited institutional fact-check; the Misinformation Severity Score is an
evaluative conclusion about communicative intent, not a statement of criminal guilt or a
finding that the speaker committed a crime. The factual predicates are cited to primary
sources and institutional fact-checkers; the evaluative conclusions are protected
expression. *New York Times Co. v. Sullivan* and its progeny apply to this assessment of a
public official's public communications.

---

Section 8: Output Format (Profile Integration)

The SMA block is placed after “Overall Veracity Track Record {#track-record}” and before “Social Media Accounts {#social-media}” in the spokesperson profile template. It mirrors the DMA block structure in full.

Format A: Profile With SMA Designation

---

## Spokesperson Misinformation Assessment

**Cumulative Designation:** [Pattern of Strategic Misinformation / Sustained Disinformation Campaign]

**Comparative Baseline:** [within-norm / elevated / outlier] versus the cross-administration
press-secretary reference class. [One sentence with the raw-count basis. Carries the
"signal, not proof" and selection-bias disclosures.]

| Metric | Value |
|--------|-------|
| Qualifying statements scored | [count] |
| Highest individual MSS | [score] — [Label] |
| Primary distinguishing factors | [list of applicable misinformation categories] |
| Comparative baseline band | [within-norm / elevated / outlier] |

### Scored Statement 1: [Short title]

**Distinguishing Factor(s):** [misinformation category/categories]
**MSS:** [score] — [Label]

**Statement:** "[Exact quote]" ([date], [context — briefing / X / TV hit])

**Fact-check source:** [Organization, publication date, rating]

**Advocacy vs. Misinformation Determination:** **Misinformation.** [Explanation citing the
specific distinguishing factors met. State the legitimate-advocacy version of the message
in one sentence.]

---

### Scored Statement 2: [Short title]

**Distinguishing Factor(s):** [category]
**MSS:** [score] — [Label]

**Statement:** "[Exact quote]" ([date], [context])

**Fact-check source:** [Organization, date, rating]

**Advocacy vs. Misinformation Determination:** **Misinformation.** [Explanation]

---

### What Is NOT Scored

[Statements excluded and why: pre/post-tenure commentary (Tenure-Window Guard), advocacy
safe-harbor framing, isolated corrected errors, opinion, and any statement lacking an
institutional fact-check citation.]

### Assessment Basis

This assessment evaluates in-role, fact-checked public statements only. The tenure window
assessed is [subject_tenure_start] to [subject_tenure_end]. Each scored statement cites its
institutional fact-check and knowledge basis. The comparative baseline is an ordinal band,
not a rate. [Add any profile-specific notes.]

### Legal Disclaimer

The Spokesperson Misinformation Assessment is an analytical framework applying defined
criteria to documented, in-role public statements by a government spokesperson. Each score
rests on a cited institutional fact-check; the Misinformation Severity Score is an
evaluative conclusion about communicative intent, not a statement of criminal guilt or a
finding that the speaker committed a crime. The factual predicates are cited to primary
sources and institutional fact-checkers; the evaluative conclusions are protected
expression. *New York Times Co. v. Sullivan* and its progeny apply to this assessment of a
public official's public communications.

Format B: Profile With No SMA Designation

---

## Spokesperson Misinformation Assessment

**Designation:** None

**Advocacy vs. Misinformation determination:** This spokesperson's documented in-role
statements represent [favorable framing of contested policy / ordinary message discipline /
isolated errors] pursued through legitimate advocacy. Where statements were rated false by
institutional fact-checkers, the record does not [meet the 3-statement threshold / show
repetition after correction / establish knowledge] required for an SMA designation.

**Comparative Baseline:** [within-norm / insufficient in-window fact-check record] versus
the cross-administration press-secretary reference class.

**Framework note:** The Spokesperson Misinformation Assessment evaluates the transition from
advocacy to knowing misinformation, not one-sided spin. Vigorous advocacy for an
administration's agenda — even aggressive or misleading-by-emphasis — is protected message
discipline, not misinformation.

**Framework disclosure:** This assessment applies a published analytical framework to
documented, in-role public statements by a public official, filtered to their tenure window.
All factual predicates are cited to institutional fact-checkers. This assessment is subject
to update as new evidence emerges or prior evidence is corrected.

---

Section 9: Anti-Patterns and Guardrails

What This Skill Is NOT

• NOT an advocacy scorer — vigorous, one-sided messaging for the President’s agenda is never misinformation on its own.
• NOT a veracity re-litigator — it consumes institutional fact-check ratings; it does not invent its own.
• NOT a tone judge — combative or hostile delivery, absent a false factual claim, earns no score.
• NOT applicable to private-citizen speech — pre/post-tenure TV commentary is categorically excluded.
• NOT a substitute for the per-statement veracity ratings — it aggregates them; spokesperson-profile-builder produces them.

Hard Guardrails

1. Tenure-Window Guard runs FIRST — no scoring of out-of-window or out-of-capacity speech.
2. Gate runs before scoring — no MSS without an Advocacy-vs-Misinformation classification.
3. Institutional fact-check required — no fact-check citation, not eligible; outcome inference alone caps at MSS 1.
4. Knowledge must be documented for MSS 3+ — a “False” rating alone caps at MSS 1.
5. Bipartisan baseline — the reference class must include both parties’ spokespeople.
6. Ordinal band, not a rate — never publish a computed “lie rate”; cite raw counts.
7. Advocacy Safe Harbor is absolute — favorable framing of true facts is never scored.
8. Sanity-check composition — patriot-sanity-check runs last and may flag disproportionate scores.
9. Transparency is mandatory — every classification includes reader-facing reasoning and the legitimate-advocacy alternative.

Common Mistakes to Avoid

Mistake	Correction
Scoring post-tenure Fox/CNN commentary	Categorically excluded by the Tenure-Window Guard
Scoring aggressive spin as misinformation	Favorable framing of true facts is advocacy, not misinformation
Inferring knowledge from a false rating	A “False” rating alone caps at MSS 1; need knowledge evidence for higher
Publishing a “lie rate” percentage	Use the ordinal baseline band; cite raw counts; disclose selection bias
Re-rating a statement’s veracity	Consume the institutional fact-check; do not substitute your own rating
Scoring a one-off corrected error	Isolated, corrected, non-repeated errors are safe-harbor advocacy
Baselining against one party only	The reference class must be bipartisan

---

Section 10: Composition with Other Skills

The SMA sits in a fixed position within the spokesperson profile skill chain:

1. spokesperson-profile-builder          (writes statements + per-statement veracity ratings)
2. accountability-profile-verification    (evidence and sourcing)
3. spokesperson-misinformation-assessment (SMA scoring — THIS SKILL)
4. patriot-sanity-check                   (proportionality validation — runs last)

Interactions

• Receives from spokesperson-profile-builder: the Significant Public Statements (exact quotes, dates, contexts) and their institutional veracity ratings — the SMA’s evidentiary input.
• Receives from accountability-profile-verification: evidence-tier classifications — determines scoring eligibility and the MSS cap.
• Reads: spokesperson-conduct-baseline.md — the bipartisan reference class and ordinal bands for the comparative layer.
• Passes to patriot-sanity-check: the SMA block for proportionality review.
• References: malice-evaluator (gate-then-score and defamation-framework precedent), international-mining-deal-specialist (deviation-from-baseline precedent), patriot-speech-analyzer (rhetoric analysis, distinct from veracity scoring).

---

Section 11: Examples

Example A: Statement Classified as MISINFORMATION (MSS 3)

Subject: A sitting White House press secretary (in-window).

Statement: “This bill does not add to the deficit,” restated after the Congressional Budget Office published a $3.8T deficit projection.

Tenure-Window Guard: In-window (sitting press secretary). Eligible.

Advocacy vs. Misinformation Gate:

• Boundary test: could this be favorable spin? A claim that a bill is “fiscally responsible” would be advocacy. But asserting it “does not add to the deficit” contradicts the nonpartisan CBO score that was public at the time.
• Factors met: Internal contradiction (contradicts the government’s own scorekeeper — 3), and the claim was repeated after the score was public (Debunked repetition — 1). Institutional-trust adjacency (dismissing CBO). Classification: Misinformation.

Category: Internal Contradiction; Debunked Repetition. Fact-check source: PolitiFact, May 21, 2025, “False”; FactCheck.org, May 2025. MSS: 3 (Knowing Distortion) — the CBO score, the government’s own authoritative fiscal estimate, was public before the restatement; a press secretary is charged with knowledge of it.

Transparency output:

Advocacy vs. Misinformation Determination: Misinformation. The statement contradicted the CBO’s published $3.8T deficit projection — the government’s own nonpartisan scorekeeper — and was restated after the score was public. The legitimate-advocacy version of this message would be: “We dispute the CBO’s assumptions and believe the bill’s growth effects will offset its costs.”

Example B: Statement Classified as ADVOCACY (No Score)

Subject: An agency spokesperson (in-window).

Statement: “This is the most transparent administration in history.”

Advocacy vs. Misinformation Gate:

• Boundary test: this is a contestable value claim / puffery, not a discrete factual assertion an institutional fact-checker has rated false.
• No factor met; no institutional fact-check citation. Classification: Advocacy.

Output: No SMA score. Noted in “What Is NOT Scored” as safe-harbor puffery.

Example C: Statement Excluded by the Tenure-Window Guard

Subject: A former press secretary, now a cable-news contributor.

Statement: A false claim made on a TV panel two years after leaving the podium.

Tenure-Window Guard: Out-of-window (post-tenure). Excluded — private-commentator speech, not government messaging. Not scored; noted once in “What Is NOT Scored.” (This person may still appear as a data row in spokesperson-conduct-baseline.md for their in-office record only.)