# Claims Clearance

> Claims Clearance

- Skill: `everyinc/claims-clearance` (Agent Skill)
- Install (CLI): `npx skillmds@latest add everyinc/claims-clearance`
- Raw SKILL.md: https://api.skillmd.com/api/skills/everyinc/claims-clearance/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: everyinc (https://skillmd.com/u/everyinc)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/everyinc/claims-clearance

---

# Claims Clearance

The pre-ship verification gate for every launch surface. Inventories every factual claim
on every asset — stats, quotes, logos, prices, availability statements, competitor
comparisons — and clears, flags, or blocks each one.

## When to invoke

- Week 8 of every L/XL launch, before anything ships (canon Week 8 checklist item)
- Any copy round review ("check the claims in this," "is this stat sourced")
- `codex-gut-check` hands over its unverified-claims inventory
- A price, stat, quote, or logo appears in any draft headed for a public surface

## The clearance categories

Every claim lands in exactly one:

- **CLEARED** — source cited inline (live link or named document + date), or permission
  logged (who granted, when, in writing where).
- **CONFLICTED** — two launch surfaces disagree. Conflicts block both surfaces until the
  owner rules; for pricing, that is Brandon and the product GM.
- **UNVERIFIED** — no source. Unverified is not a status that ships; the claim is cut or
  cleared, never softened into shipping ("many users say…" is the tell for laundering an
  unverified stat).
- **PENDING PERMISSION** — logos, testimonial quotes, named-person or named-company
  references awaiting a written yes. No logo ships without one.

## The sweep

1. Inventory every surface in the burn-down: site, emails, social assets, video
   end-cards, pricing page, press materials, changelog.
2. Extract every claim — numbers, superlatives, comparisons, names, logos, prices,
   availability ("works with X," "the only Y that Z").
3. Check each the same way. No sampling. If the sweep runs out of room, name what wasn't
   covered rather than implying it passed.
4. Cross-surface consistency pass: the same fact must read identically everywhere
   (price, plan names, trial length, stat wording).
5. Deliver the clearance table (claim · surface · category · source/owner · action) and
   a plain block list of what cannot ship as-is.

## Rules

- The bar for "verified" is fixed before the sweep and doesn't move; narrowing a
  definition to keep a claim alive is the tell for rationalizing.
- Comparative claims about competitors require a live check dated within the launch
  window (`marketing-science/research/references/competitive-audits.md`), not memory.
- Clearance is recorded in the burn-down task's note, so the gate is auditable at the
  final design approval.

## Dependencies

- `foundation/marketing-os`
- `launches/canonical-process.md` (epistemic rules, Week 8 placement, pricing ruling)
- `marketing-science/research` (live re-verification of comparative claims)

## Quick checklist

- [ ] Every surface inventoried; every claim extracted
- [ ] Each claim in exactly one category; same check for every item
- [ ] Conflicts routed to their owner (pricing → Brandon + GM)
- [ ] Unverified claims cut or cleared — never reworded into shipping
- [ ] Permissions logged: who, when, where in writing
- [ ] Cross-surface consistency pass done
- [ ] Clearance table + block list delivered; uncovered items named