brand-safety-complianc…
Scans outbound copy or creator-submitted content for 6 compliance red lines — medical/efficacy, exaggeration/absolute, beauty-permanence, financial/income,…
Maps a creator's evidence to a collaboration tier (Open / T1 / T2 / T3·VIP) and judges whether they hit the Open→Target graduation line. Use when the user needs to decide which tier a creator belongs to, whether to graduate them, or what priority to give them; also called by
$ npx -y skills add crevideo/crevideo-reach --skill creator-tier-resolver --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/creator-tier-resolverContext preview
The summary Claude sees to decide when to auto-load this skill.
Maps a creator's evidence to a collaboration tier (Open / T1 / T2 / T3·VIP) and judges whether they hit the Open→Target graduation line. Use when the user needs to decide which tier a creator belongs to, whether to graduate them, or what priority to give them; also called by
name: creator-tier-resolver description: >- Maps a creator's evidence to a collaboration tier (Open / T1 / T2 / T3·VIP) and judges whether they hit the Open→Target graduation line. Use when the user needs to decide which tier a creator belongs to, whether to graduate them, or what priority to give them; also called by creator-fit-scoring and offer-policy-checker. Affiliate (分销) context. Chinese triggers: 判定分级 / 该不该毕业 / 用什么优先级对待.
**Role**: answer only "which tier this creator should be in, and whether they hit the graduation line". **No scoring** (that's creator-fit-scoring), **no commission validation** (that's offer-policy-checker), **no copy**.
Write every output in the **merchant's working language**, using that market's native seller terminology:
Tool names stay identical in both languages. If unsure which market, ask once before producing output.
1. Creator evidence: conversion / GMV-per-view, orders, fulfillment, audience persona; follower count (reference only). 2. Optional: target product category (category affects tier economics — see the Key-Rules Cheat Sheet).
1. **Hard-exclude check**: blocklist / explicit opt-out / risk-signal hit → "Excluded / do not invite". 2. **Conversion evidence is primary**: judge the tier against the Cheat Sheet's tier definitions. **Follower count is not the primary basis**; category-sensitive (the same numbers may land in a different tier per category, per the registry). 3. **Insufficient data or on a tier boundary**: take the **lower** tier (conservative) and flag "insufficient/borderline evidence, recommend human confirm or more data". 4. **Graduation check**: per the Cheat Sheet's graduation line (last-N-day GMV or count of converting videos), judge whether an Open creator should rise to Target; if they qualify → mark "recommend graduation + go via Target Collaboration invite". 5. Output tier + basis + next step (hand to offer-policy-checker for the offer, outreach-message-composer for the draft).
This skill embeds **no** tier thresholds, graduation line, commission numbers, or tool list. Tier definitions, the graduation line, and category-sensitivity rules all come from the Action Workspace Doctrine → Key-Rules Cheat Sheet; if an example here conflicts with current policy, the Cheat Sheet wins. Only call tools that actually exist.
Render the report in the merchant's working language; the template below shows the structure.
🎚️ Tier: T2 (or Open/T1/T3·VIP/Excluded/insufficient-evidence) [Basis] key conversion evidence… · audience fit… · checked against Cheat Sheet tier defs · no hard-exclude signal [Borderline/insufficient] (if any) just short of the next tier / missing X data → recommend human confirm or more data [Graduation] (Open creators) hit the registry graduation line? → recommend graduation via Target Collaboration / not yet [Next] → offer-policy-checker for the offer; → outreach to draft (warm creators prefer Target Collaboration)
Input: a creator with steady repeat purchase, audience fit, mid followers. Output: 🎚️ Tier = **T2** (on conversion evidence, checked against the Cheat Sheet tier defs; not on followers). Did not reach a higher tier; if just short, flag "recommend more data / human confirm". Next → offer-policy-checker for the offer.
Crevideo Reach — turn Claude (Cowork / Claude Code) into your TikTok Shop affiliate outreach & marketing assistant; the affordable, agent-native alternative to Euka and Reacher.
Scans outbound copy or creator-submitted content for 6 compliance red lines — medical/efficacy, exaggeration/absolute, beauty-permanence, financial/income,…
Scores one or more TikTok Shop creators for real sell-through fit using evidence (conversion over followers), returning a fit verdict + recommendation strength…
Deep reference / playbook for correctly driving the crevideo-reach MCP — TikTok Shop affiliate (分销) and affiliate (分销) work: tool map (79 tools), the affiliate…
Validates a proposed commission/offer/outbound action against policy: within the tier's commission cap, all-in take-rate under the gate, product ACTIVATE,…
Drafts TikTok Shop creator outreach copy across three channels — DM / Target Collaboration invite / email — tuning tone by tier, localizing to brand voice,…
Reads the 3 affiliate reports (collaboration / product / shoppable-video) and produces decision-grade insight itself — key metrics + 5 leading indicators + a…