# Viral Claim Auditor

**Trend2Truth EXP-002 product pack**  
Price test: free now → $19 later  
Not financial advice.

Use this when a creator promises monthly income from a “simple” system. The goal is not vibes — it is a go / no-go for a controlled experiment.

---

## Prompt 1 — Strip the ad

Paste the caption, script, or landing-page copy, then ask:

```
You are an auditor, not a hype coach.
1. Restate the income claim in one sentence with numbers and timeframe.
2. List every action the viewer must perform.
3. List every asset they must already have (audience, capital, skills, legal entity).
4. Separate METHOD from MONETIZATION HOOK (course, discord, coaching, software).
5. Flag unverifiable assertions.
Return markdown with those five headings only.
```

---

## Prompt 2 — Cost & time model

```
Using the cleaned claim, build a 30-day experiment plan:
- Weekly hours (realistic, not influencer hours)
- Cash costs (tools, ads, inventory, fees)
- Leading metrics that show up before revenue
- Hard stop rules (when to quit)
- Human gates an AI cannot pass (KYC, tax, platform identity)
Assume the operator is a solo AI agent plus one human for identity tasks only.
```

---

## Prompt 3 — Proof demand list

```
Write a proof checklist the claimant should be able to show:
- Primary source revenue evidence
- Traffic or distribution evidence
- Time-to-first-dollar narrative with dates
- Failure modes they hit
If they only show lifestyle footage, grade proof as FAIL and explain why.
```

---

## Scoring rubric (0–20 each)

| Dimension | 0 | 10 | 20 |
|---|---|---|---|
| Specificity | Vague niche + vague money | Named channel, fuzzy math | Clear offer, niche, unit economics |
| Effort honesty | “Passive / no work” | Some work admitted | Labor and skills named |
| Proof | Screenshots of nowhere | Partial dashboards | Verifiable primary sources |
| Risk clarity | Hidden downside | Soft caveats | Explicit risks + kill criteria |
| AI-operability | Needs body/vehicle/storefront | Mixed | Mostly digital work an agent can ship |

**Total /100**

- **0–39:** Marketing costume. Do not start.
- **40–69:** Narrow it before spending.
- **70–100:** Eligible for a Trend2Truth-style public test.

---

## One-page experiment brief (copy/paste)

```
Title:
Claim under test:
Hypothesis:
Operator (AI / human split):
Budget cap:
Time cap:
Leading metrics:
Revenue definition:
Human gates:
Start date:
End date / verdict date:
Journal URL:
```

---

## Example tear-down seed

Claim: “Faceless AI YouTube pays $10k/month.”

Expected auditor notes (abridged):

- Method: script → edit → upload → watch ads.
- Monetization hook: often a editing pack / community.
- Human gates: YouTube identity, AdSense, copyright strikes, thumbnail taste tests.
- Leading metrics before $10k: watch time, CTR, RPM by niche — not subscriber vanity.
- Verdict posture: test script throughput first; treat revenue claims as unverified until RPM math is public.

---

Built by Trend2Truth for EXP-002.  
https://github.com/l28bit/Trend2Truth