# Creatorland Data MCP, primer for your agent

Paste this block into your AI client's custom instructions, project knowledge, a
CLAUDE.md, or an AGENTS.md so your agent self-orients on what the Creatorland
Data MCP can do, what it costs, and the rules of the road. Everything below is
public and safe to share with an agent. Safety is enforced by the server; this
primer steers good usage, it is not the gate.

## What it is

The Creatorland Data MCP is a connector that puts Creatorland's first-party
creator and deal data inside your AI client. With it your agent can find and rank
creators, read a creator's profile, benchmark deal rates, see which brand
categories are active for a creator, run bulk match-and-enrich over a contact
list, pull a verified audience report (real demographics, geo, and a
fake-follower score), and (on paid plans) start consented brand-to-creator
outreach. You stay in
your AI client the whole time; the MCP is the data layer behind it.

Connect URL (MCP endpoint): https://mcp.creatorland.com/mcp
Auth: OAuth 2.1. You are only charged for successful calls that return data.

## Tools and credit cost

- search_creators (2 credits): find and rank creators two ways. Brief mode takes
  a natural-language campaign description; lookalike mode takes a seed creator.
  Hard filters: geo or region or state, platform, follower tier, niche,
  freshness, brand affiliations.
- get_creator_profile (1 credit): one creator's identity, platforms and follower
  counts, location, niche and interests, freshness, and (paid) brand
  affiliations and avatar.
- query_market_intelligence (5 credits): market aggregates: deal volume, rate
  bands (p25, median, p75), deal-type mix, and recency for a vertical or tier.
- get_member_content_stats (1 credit): account-level, OAuth-consented content
  stats for a live member.
- match_creators (free): bulk-resolve a list of identifiers or handles to known
  creators; returns counts and a credit estimate. Abuse-bounded.
- enrich_matches (3 credits per funded match; Pro and pilot): append interests,
  geo, freshness, and affiliations to matched creators in bulk, against a cost
  you confirm first.
- request_creator_connection (10 credits; Pro and pilot): consented
  brand-to-creator outreach. One charge runs the whole sequence plus reply
  monitoring and intro brokering.
- get_connection_status, list_connections (free): track outreach requests.
- check_audience_coverage (free): for a list of creators, how many already have
  a cached audience report vs need a fresh paid pull, with a credit estimate
  before you spend.
- get_audience_report (25 credits; Pro): a verified audience report for one
  creator: real demographics, geo, interests, audience brand-affinity, and a
  credibility / fake-follower score. Instagram is fullest; TikTok and YouTube
  are demographics + geo only. Cache-first, so a creator pulled in the last ~30
  days re-serves free.
- health (free): liveness and data-store reachability.

## Plans

- Free: 250 credits a month, no card. Every read tool (search, profile, market
  intelligence, brand affiliations, member content stats), bounded by the credit
  grant. Out of credits returns a clean upgrade message, not an error.
- Pro: 2,000 credits a month plus the paid tools (outreach, avatars, bulk
  enrich, audience reports). Credit packs at $25 per 1,000 cover overage, with
  volume discounts; packs never expire.
- Pilot: contracted agencies, unlimited.

A full workflow is roughly 10 to 30 credits, so the free grant is about 15 to 25
real workflows a month. Full pricing: https://mcp.creatorland.com/docs/pricing

## Skills (packaged workflows)

Install in a plugin-capable client: /plugin marketplace add creatorland/creatorland-mcp-skills

- Casting: brief-to-shortlist, transcript-to-shortlist, cast-from-a-vibe,
  lookalike-ladder, triangulated-casting, longlist-machine, casting-gap-analysis,
  cross-market-slate, zero-brief-discovery, draft-day.
- Pricing: one-number-rate, fair-price-brief, rate-card-generator,
  rate-justifier, budget-allocator.
- Brand and outreach prep: creator-brand-matchmaker, competitor-watch,
  conflict-check, creator-diligence-report.
- Market and proposals: vertical-briefing, vertical-forecast-brief, deck-inserts,
  rfp-response-builder, campaign-package-builder, brief-builder,
  wrap-report-skeleton.
- Roster and data ops: roster-enricher, talent-scout, search-memo,
  skeptics-second-pass, credit-budget-guardian, onboarding-tour.

## Playbooks (goal to workflow)

- Cast a campaign: search_creators (brief) then get_creator_profile on the top
  picks, optionally conflict-check. Skill: brief-to-shortlist.
- Find more like one creator: search_creators (lookalike) then tier by price.
  Skill: lookalike-ladder.
- What should I pay: query_market_intelligence for the vertical and tier. Skills:
  one-number-rate (fast) or fair-price-brief (memo).
- Where should this creator pitch: get_creator_profile then
  query_market_intelligence across their categories. Skill:
  creator-brand-matchmaker.
- Same campaign in several countries: geo-constrained search_creators per market.
  Skill: cross-market-slate.
- Build it into my app: wire the tools directly and keep this primer in context.

## Rules of the road

- Every aggregate carries provenance (what it is derived from).
- A minimum-N privacy floor withholds or broadens slices that would describe too
  few creators or deals; broadening is disclosed.
- No contact information is ever exposed. Outreach happens only through the
  consented connection flow.
- Deal-activity patterns are not brand intent. An active category is not a warm
  lead.
- Sample outputs and rate bands are illustrative market-level figures, not a
  specific creator's price or a guarantee.
- Paid capabilities are refused with a clean upgrade message when your plan lacks
  them; they are never silently widened.

Use cases by audience: https://mcp.creatorland.com/use-cases