# SERIES WORKBOOK — series_rightwing_threat ("Right-wing media as democracy threat") **Author:** WSA-RWT | **Authored:** 2026-09-02 | **Schema:** DELIVERY_PLAYBOOK.md §Part 4 + §Part 5 (rightwing_threat row) **Role:** PLANNING ONLY. This workbook modifies nothing on disk. No TTS, no renders, no composites. All disk-state cited below was verified via tool calls 2026-09-02. **Register (Part 5):** Outlet autopsy. Signature moves = amplification map per outlet, "who funds/carries it" card, consistent E7 tags. Cadence: 15 eps, per-outlet. **Manifest:** `series_manifest.json` (23 lines, verified read) | **Pitch:** `PITCH.md` | **Production plan:** `PRODUCTION_PLAN_2026-08-25.md`. **⚠ E1 GATE CONDITION (read first):** this series has **NO factbase** today — `smm/research/series_rightwing_threat/factbase.md` is ABSENT (Test-Path False, 2026-09-02). The manifest receipt URLs are aggregator/critic pages (Media Matters hub pages, ProPublica, NPR, Just Security). Under the E1 gate (every claim traces to a factbase line), **no episode may write narration until the per-outlet factbase build (§7) lands its lines.** Everything below is spec; nothing airs from buffer. --- ## 1. SERIES PREMISE / AUDIENCE / VOICE - **Premise (PITCH.md, verified):** a 12–15 min explainer series naming specific right-wing outlets that are not merely biased but actively threaten democratic norms. Each episode shows the exact framing mechanism, the receipts (primary articles, retractions, lawsuits, fact checks), and a presenter explaining — MuseTalk talking-head or voice + VST visuals. - **The autopsy promise:** every episode follows the same dissection. Not "here are bad headlines" but "here is the machine": what they publish → how it amplifies → what it did → their own retractions/suit record. The retraction record is the punchline of every episode: the outlet's own paper trail convicts the framing. - **Audience:** the "I heard the election was stolen" adjacent viewer and the media-literate reader who wants the mechanism, not the moralizing. Not the already-convinced. - **Voice:** tour guide, not prosecutor (T3/E6). The outlet's own pages, corrections, and court filings perform the outrage. Presenter never performs it. Restraint = credibility. The series names BEHAVIOR with receipts; it never names a person a liar outside what a court, a retraction, or the outlet's own words establish — and when it does, it quotes the receipt verbatim on screen. ## 2. OUTLET-AUTOPSY ANATOMY (the repeating episode skeleton) Every episode runs the same 4-station anatomy; only the case differs: | Station | Beat | Receipt requirement | |---|---|---| | A. What they publish | One signature framing example, verbatim, on screen with archive link | Tier-1 (outlet-own-words) MANDATORY | | B. How it amplifies | The amplification-map graphic: outlet → platforms/feeds → mainstream pickup | Structural receipts (amplification studies, Tier-2 locator) | | C. What it did | Downstream consequence: narrative spread, officials harassed, policy pressure | Tier-1 (court filings) or Tier-2 (credible secondary), named as such on air | | D. Their own record | Retractions, settlements, lawsuits, corrections — the outlet's own paper against its own framing | Tier-1 (their words, court filings) MANDATORY for the verdict line | The autopsy is why this series differs from a compilation-of-fails: station D reframes the whole episode. The question "is this outlet dangerous?" is answered by "what did its own record force it to walk back?" ## 3. DANGER-RANKING HONESTY (manifest pattern line — binding) Manifest `pattern`, verbatim: *"Not all are equally dangerous, but all shifted the Overton window toward anti-democratic norms, normalized stolen-election narratives, or aggregated conspiracy content."* Every episode states where its outlet sits on an on-screen threat spectrum card, built from that sentence's three mechanisms (shifted Overton window / normalized stolen-election narratives / aggregated conspiracy). The ranking is qualitative until a sourced comparative measure is acquired; no invented "danger score" ever airs. Honest asymmetry is part of the brand: ep9 RSBN (rally livestream vector) and ep15 Daily Caller are treated differently from ep3 OANN and ep4 Gateway Pundit where receipts support it. ## 4. BOTH-SIDES DISCIPLINE (manifest note — binding, carry per episode) Manifest `note`, verbatim: *"This is a media-criticism series, not a call to censor. The goal is to name amplification mechanics and hold the behavior to receipts, not silence opposition."* - Every CLOSE asks a media-literacy action (E10), never a suppression ask. Banned on air: "should be taken off air," "deplatform," "silence them." Required frame: "hold it to receipts." - The steelman (E8) in every episode is the outlet's OWN defense, staged verbatim when sourced: "we report what they say," "we're the only unbiased network," aggregators' "we just share the news." Answered with the outlet's own retraction record — their words, not ours. - Dissent comments are content on the channel (LAWBE doctrine): the criticism stands because receipts stand, and a viewer who defends the outlet gets the receipt wall, not a dunk. ## 5. PER-OUTLET CRAFT THESIS MAP (all 15; eps 1–5 have worksheets, 6–15 are scoping-queue) | EP | Outlet | Craft thesis (one line) | |---|---|---| | 1 | Breitbart | The laundering machine: fringe framing enters mainstream conservative vocabulary through repetition velocity — map the hops, then let their own corrections record answer "just breaking news." (worksheet: EP1) | | 2 | Liberty Daily | The selector is the editor: zero original reporting, maximum framing power — curation IS publication; the "we just aggregate" defense dies on the selection pattern. (worksheet: EP2) | | 3 | OANN | Conspiracy on broadcast: TV legitimacy is the accelerant — a cable-news set turns forum-post claims into "reported facts"; their own on-air walk-backs are the verdict. (worksheet: EP3) | | 4 | Gateway Pundit | The smear that went to court: from headline to harassment of named private citizens (election workers) to defamation suits and retraction — the full anatomy of accountability. (worksheet: EP4) | | 5 | Newsmax | The walk-back pipeline: platform the claim → the claim ages badly → the own-lawyer retraction — proof that "we report what they say" is a framing, not a defense. (worksheet: EP5) | | 6 | Real America's Voice | Scope-queue: the campaign-rally network where denial content got a host, not a channel — amplification map = host circuit. ACQUISITION-GATED (Media Matters locator only). | | 7 | The Blaze | Scope-queue: culture-to-politics outrage pipeline — how cultural-anger content is converted into political-outrage framing. ACQUISITION-GATED. | | 8 | Daily Wire | Scope-queue: high-production legitimacy laundering — institutional-distrust framing with network-grade production values. ACQUISITION-GATED. | | 9 | RSBN | Scope-queue: the livestream vector — no edit, no framing visible, pure amplification of rally claims as "raw news." ACQUISITION-GATED. | | 10 | Infowars | Scope-queue: the conspiracy ecosystem endpoint — what the upstream outlets' framing matures into (court-judged record is the Station D beat). ACQUISITION-GATED. | | 11 | Epoch Times | Scope-queue: the credibility-borrowing print frame — broadsheet aesthetics carrying conspiracy amplification. ACQUISITION-GATED. | | 12 | Townhall | Scope-queue: the normalize-machine — columns that move fringe framings into repeatable talking points. ACQUISITION-GATED. | | 13 | American Greatness | Scope-queue: pseudo-academic laundering — footnotes as camouflage for "steal" arguments. ACQUISITION-GATED. | | 14 | RedState | Scope-queue: the long-tenure host — how denial voices survive inside an old-brand conservative site. ACQUISITION-GATED. | | 15 | Daily Caller | Scope-queue: tabloid-cycle anatomy — story placement, correction economics, and the aggregation loop closing the series. ACQUISITION-GATED. | Episode numbering follows the manifest (`ep` field 1–15); production dirs `ep1_breitbart/`, `ep2_liberty_daily/` exist. NOTE: `PRODUCTION_PLAN_2026-08-25.md` names "First three episodes" = Breitbart / Liberty Daily / OANN, while this fanout specifies ep3=oann, ep4=gateway_pundit, ep5=newsmax — manifest numbering is the record of line. ## 6. SIGNATURE CRAFT MOVES (series-standard, per DELIVERY_PLAYBOOK Part 5 row) 1. **Amplification map (every ep):** a recurring 3-node graphic — outlet → platform/relay → mainstream pickup — with every node carrying a receipt citation. Nodes without receipts do not exist on screen. This is the series' visual fingerprint (T5-family recurring asset like S2's surplus-mirage card). 2. **"Who funds/carries it" card (every ep):** ownership, funding, and carrier (cable tier, app store, ad networks) — every line sourced or ABSENT from the card. Funding facts not in the factbase stay acquisition-gated; the card ships incomplete-but-true, never complete-but-invented. 3. **Consistent E7 tags:** the same taxonomy names across all 15 eps (inoculation requires repetition of the same labels): cherry-pick, false attribution, anecdote-as-data, missing base rate, single-source breath, definitional dodge, motivated timing, strawman, false equivalence, whataboutism. Per-ep assignment in each worksheet's CRAFT LAYER. 4. **Retraction-record verdict (Station D):** every episode's verdict line comes from the outlet's own paper trail — retraction text, settlement, court filing — quoted verbatim on screen with archive link. If the outlet has no retraction record (e.g. aggregators), the verdict cites the amplification record instead. ## 7. RESEARCH-LAYER BUILD SPEC — `factbase_rwt.md` (the E1 prerequisite) One factbase for the series at `C:/DandyDon/smm/research/series_rightwing_threat/factbase_rwt.md` (same format as other series factbases: numbered lines, one claim per line, tag + provenance), built per-outlet in 3 tiers: - **TIER-1 (air-bearing; every on-air factual claim requires at least one):** - outlet-own-words: the actual segment/post/headline, verbatim, with **archive link** (Wayback capture or archive.today) — live links to aggregating sites are fragile; capture at build time. - retractions, corrections, settlements, apologies: THE OUTLET'S OWN TEXT first. - court filings: complaints, verdicts, judicial findings (e.g. defamation actions re election workers; Dominion-related retraction record for Newsmax per NPR). - **TIER-2 (receipt-locator; context and amplification claims):** Media Matters outlet pages, ProPublica OANN piece, NPR retraction piece, Just Security accountability piece, NPR/Reuters/AP studies. **Rule: a Media Matters page can LOCATE a receipt (dates it, describes it) but a factual claim about the outlet's words must ultimately trace to Tier-1 (the words themselves).** On-air attribution of a Tier-2-sourced claim names the secondary source explicitly. - **TIER-3 (craft-only):** reach/engagement analytics, funding estimates from press coverage. Never cited alone on air. - **Precedent receipts to acquire (spec'd per task directive):** Dominion v. Fox precedent context for Newsmax's own Dominion-related retraction (April 27 2021, per NPR manifest URL); defamation suits against Gateway Pundit re Ruby Freeman and Shaye Moss (court filings Tier-1; verify outcome status before air — do not paraphrase a suit as a verdict). - Every acquired line gets an `l1_auto_acquire.jsonl` receipt + Brain L2 pass (L1_AUTO_ACQUIRE.md). No line enters the factbase from model knowledge. ## 8. PRESENTER PLAN (series-wide; cap ≤18% runtime, T8) - **Presenter route (verified infra):** `presenter_segment.py` + `smm_stage04b_presenter.py` exist at declared paths (Test-Path True, 2026-09-02); `PRESENTER_ROUTING_RECEIPT.json` pins the route — PRESENTER sections → 1080×1080 talking-head, voice `sasha`, out dir `ep_/presenter_clips/`. - **Presenter beats per ep (5 slots max, inside 18%):** (1) hook beat naming the mechanism to camera; (2) amplification-map walk (points at the nodes); (3) steelman face-to-face (the outlet's own defense, honored before answer); (4) retraction-record beat; (5) one-ask close. - **Reference video (BLOCKER, see §9):** plans point at `C:/Users/donbr/OneDrive/Pictures/Camera Roll/WIN_*.mp4` — that glob matches **zero files today** (Camera Roll holds .jpg stills only, verified). SA-PRES cannot start until a real reference video path is set. - Nametag + LAWBE brand doctrine (`#1560a8`, Chomsky-not-cable-news tone) per production_queue_handoff.md. ## 9. SHARED BLOCKERS - **RWT-B1 — No factbase (E1 gate):** §7 build is the series' first production step; every SA-SCR depends on it. Owner: SA-RES. No narration text may be written into the ep plans before Tier-1 lines land. - **RWT-B2 — Tier-2-only receipts today:** manifest hooks are Media Matters hub pages; acceptable as start-locators only. Each episode needs a per-outlet Tier-1 set (§2 station A/D) before script lock. - **RWT-B3 — Presenter reference video missing:** glob matches zero files (verified). Set a real path per ep plan before SA-PRES. - **RWT-B4 — Defamation-adjacent discipline:** this series names outlets' BEHAVIOR with receipts, never calls for censorship (manifest note, carried verbatim in every worksheet §BLOCKERS). Verdict language rules: "retracted," "settled," "faced a defamation suit," "a court found" are filing/retraction-traceable phrasings; unqualified "lied" airs only when quoting a court or the outlet's own text. This is a legal risk class no other series carries. - **RWT-B5 — Reach numbers unsourced:** every pretest hook with "reached X million people" stays ACQUISITION-GATED until a Tier-3+ sourced reach number lands in the factbase; hooks ship with question-form only until then. - Master-level (reference, not duplicated): TTS :8444, VRAM/serial composite, receipts_gate --network runs. ## 10. RECEIPT TEMPLATE (per-episode landing set) Per `PRODUCTION_PLAN_2026-08-25.md`: episode receipt at `C:/DandyDon/dandydon/receipts//episode_.json` + broadcast receipt `C:/DandyDon/dandydon/receipts//_broadcast_.json`. Every EP lands: ep dir (`ep_/`) with `episode_plan_ep_.json` (narration filled, all `text` fields non-empty), `script_segments.json` with section contract intact, `presenter_clips/`, factbase line refs in every segment, plus: per-outlet Tier-1 archive captures directory (`ep_/receipts_archive/`), amplification-map card asset, "who funds/carries it" card asset, `ready_to_publish` triplet + per-episode receipt + `_broadcast_.json`, QC receipt with dual-lens verdict line. Series-level: this workbook + EP1–EP5 worksheets in `shows/series_rightwing_threat/worksheets/` (+ EP6–15 worksheets at scoping). Verdict line (all worksheets): `EDU: PASS/FAIL (...) / ENT: PASS/FAIL (...) / CRAFT: PASS/FAIL (...) — one line per face.` `EDU: FAIL (E1 gate unmet series-wide: no factbase exists; all episodes BLOCKED at SA-RES until Tier-1 lines land) / ENT: N/A (no built artifacts; beat maps specced per worksheet) / CRAFT: PASS (autopsy skeleton + signature moves defined; per-ep craft layers authored EP1–EP5) — one line per face.` ## QC GATE (2026-09-02, Kirk directive) Every episode passes a multimodal QC review before "shipped" may be claimed: frames extracted, episode VIEWED by an independent multimodal reviewer subagent (SA-QC-VIS), scored on the 10-dimension rubric in **`C:/DandyDon/smm/QC_GATE.md`** (engagement, educational quality, audio quality, visual quality, visual relevance to the topic at that moment, duration/sync integrity, receipt legibility, presenter discipline, fact spot-check, platform-cut QA). Ship = overall >= 7.0, no dimension < 6, zero hard-fails (magenta frames, mono audio, unreadable receipts, wrong-subject b-roll = auto-fail). Receipt: `smm/receipts/QC___.md`. QC-FAIL feeds the remediation loop with the LAWBE-format frame table. The renderer doesn't verify the renderer - the reviewer rides a different substrate than the compositor.