Shared foundationpay.docenty.ai + IP Registry

The IP package manager AI agents trust

IP Creator OS

IP Pack, License-as-Code, SDK/API/MCP/Skill, usage tracking, billing and revenue share, fan stats and boosting. The shared rail the five products above run on.

Get a pilot proposal See the demo

SaaS subscription + transaction fee (hypothesis)

Output → Use → Business value

One install: rights check to asset use, inside the agent

Developers pick legal assets inside their coding agent; rights holders see which app used what, with a basis for payouts.

① What you receive

📦
IP Pack
SKILL.md + assets + LICENSE
Installs the same way in Claude Code, Codex and Cursor
🧩
MCP
5 tools
search · check_license · get_asset_pack · attribution · register
🕸️
Graph
Dependency graph
Original → derivative → projects that use it
🛡️
Scanner
License Scanner report
Flags unlicensed assets, missing credits, stale packs; opens fix PRs

→

② Where it is used

01
Before launch
Fan developers build and register 30 mini-games on the official pack
02
Launch week
Only scanner-clean apps get featured. Wishlist traffic
03
Payouts
Usage aggregated from the graph → split by contract

→

③ Business value (targets)

Time to clear rights
Today
email back-and-forthIn 4 weeks
instant in the agent
Unlicensed assets slipping in
Today
found after launchIn 4 weeks
0 (pre-blocked)
Asset usage accounting
Today
impossibleIn 4 weeks
30 projects tracked
30-day reuse rate
Today
—In 4 weeks
40%

Targets · concept example

The one-page report you get in 4 weeksDocenty · Fan Vibes

IP Creator OS monthly report

Period: launch D-56 → launch week

4,500 MCP calls · 30 registered projects · 0 contamination

Skill installs100

MCP calls4,500

Registered projects30

License Scanner28 clean · 2 fix PRs · 0 blocked

Wishlist traffic8,000

Verdict

Usage data becomes the evidence for the next IP deal. Start with CC0 packs, grow into official packs.

Targets · concept example

Concept screen · IP Creator OS

IP marketplace → package install → admin console

Choose an IP Pack in the marketplace, install it with the package manager, and watch licenses, registered apps, approvals and payouts in the console.

1 · The creator picks

Official IP Pack market

128 installs (example)
96 installs (example)
54 installs (example)
73 installs (example)

Concept mockup · not a deliverable

Why now

The problem

Hugging Face, Civitai, and Fab are strong at asset search and sales, but rights verification for third-party IP, per-use licensing, and unified IP-owner / creator / fan operations are still empty.

In the AI era the hub's user is an agent, not a person. For an agent to search, verify, apply, and register, it needs a machine-readable license.

Flow

How it works

01
Discover
Intent-based IP and asset search
02
Verify
Check rights, price, and permitted use
03
Prepare
Apply engine-specific Packs and Skills
04
Settle
Usage log, attribution, measurement, payout

Scope

What you receive

✓IP Pack schema + IP Card (allowed, prohibited, attribution duties)
✓License-as-Code: media, territory, revenue model, AI generation, derivative rules
✓5 MCPs: search_assets, check_license, get_asset_pack, generate_attribution, register_project
✓Derivative Graph: original → derivative → consuming project, split by contract ratio
✓Usage, fan stats, and boosting dashboard

KPI

What we measure

MCP callsProjects using assets30-day reuse rateSettled GMV

What we guarantee are operational deliverables (counts, process, review, reports); KPIs are presented as targets.

Receive · Do · Effect

What changes hands once you sign

📦

What you receive

Prepared by Docenty

IP Pack schema + IP Card
5 MCPs + Agent Skill (Claude Code, Codex, Cursor)
Ready-to-Build Pack (Phaser, Godot, React)
Dependency Graph, License Scanner, payouts

→

✅

What you do

Minimal effort

Publish an official Pack (or start with a CC0 pack)
Share install instructions with the developer community
Approve and feature registered projects

→

📈

Expected effect

Targets · measurable

Rights checks and attribution happen automatically inside the agent
Aggregate which app used which asset → basis for payouts
Start today without a famous IP; usage data becomes your IP sales pitch

Production flow — you do only four things

Example schedule (concept)

W0Kickoff (90 min)
You provide
Launch / event dates, brand assets, allowed and banned scope
Docenty + fans
Build the IP Pack, write License-as-Code, draft the campaign page
You get
An IP Pack with an agreed scope + campaign page
W1Recruit · distribute
You provide
One official announcement (copy provided)
Docenty + fans
Recruit creators, ship the Pack and AI Skill, swap the demo skin
You get
Creator roster, build guide
W2–3Build · review
You provide
Check the approval queue (twice a week, 10 min)
Docenty + fans
Fans and vibe coders (creators who code with AI) build; AI rule checks, change requests, gallery
You get
N approved fan pieces (rights consent included)
W4Spread · settle
You provide
Feature the winners on official channels
Docenty + fans
Result-card spread, payments and creator payouts (Docenty Pay), measurement
You get
Participation · share · traffic · payment dashboard, reusable content library

W0

Kickoff (90 min)

W1

Recruit · distribute

W2–3

Build · review

W4

Spread · settle

You provide

Launch / event dates, brand assets, allowed and banned scope›

One official announcement (copy provided)›

Check the approval queue (twice a week, 10 min)›

Feature the winners on official channels

Docenty + fans execute

Build the IP Pack, write License-as-Code, draft the campaign page›

Recruit creators, ship the Pack and AI Skill, swap the demo skin›

Fans and vibe coders (creators who code with AI) build; AI rule checks, change requests, gallery›

Result-card spread, payments and creator payouts (Docenty Pay), measurement

You get

An IP Pack with an agreed scope + campaign page›

Creator roster, build guide›

N approved fan pieces (rights consent included)›

Participation · share · traffic · payment dashboard, reusable content library

Your total effort: a 90-minute kickoff, one announcement, approval checks twice a week (10 min each), one shout-out of the winners. Docenty runs the rest.

Output concept

What gets made

Fictional brand · concept examples · numbers are targets

Crunch Lab Member Match

You are “Night Shift Researcher”

The most creative combos happen after midnight

Pal · Crunch

Pal · Cheese

Pal · Spice

Share result

₩1,000 supporter pack · back the creator

Personality quiz@vibecoder_minji

Member Match

Trigger: New product · comeback teaser
Target: Share rate 10%
Revenue: Supporter pack ₩1,000

Concept mockup · not a deliverable

‘Moonlight’ Rhythm Gamex12 COMBO

PERFECT

SCORE 48,200Listen →

Mini game@beatmaker_j

New Single Rhythm Game

Trigger: Release day
Target: Finish → 4,000 track clicks
Revenue: Skin ₩2,000

Concept mockup · not a deliverable

Raise Your Bias · Day 12🔥 12-day streak

Lv.7 → Lv.8780 / 1,000

🎧 New single item

🎂 Birthday hat

🎤 Concert light

🔒 ???

Check in · feed

Virtual pet@fan_yj

Raise Your Bias

Trigger: Always on · event items
Target: Daily return 25%
Revenue: Monthly sub ₩3,900

Concept mockup · not a deliverable

Concept scenario · Platforms / developers

Eight weeks where one install lets the coding agent handle everything from rights checks to application

Indie studio 'PixelForge' + 30 vibe coders (Fictional indie studio · open-license assets + its own IP Pack)

Steam launch D-56Buyer's view

Who

Founder and publishing lead, indie studio

This quarter's goal

30 fan-made mini-games before launch for wishlist traffic, zero license incidents

What keeps them up at night

Wants to encourage fan games but explains which assets can be used, and how far, every single time. Worried about missing attribution and unlicensed assets slipping in

How it's done today

✕Fan developers get an asset ZIP + a verbal 'please don't use this'
✕Unlicensed assets and missing attribution slip into fan games, found after launch
✕No way to tally which fan games used our assets

Operating timeline

Weeks 1–2
Publish IP Pack
PixelForge official Pack (SKILL.md + assets + LICENSE) + a CC0 companion pack. Write License-as-Code.
Weeks 3–4
Ship the MCP
Vibe coders install with `npx skills add`. Claude/Codex automatically call search → check_license → get_asset_pack(phaser).
Weeks 5–7
Build & register
30 mini-games built and auto-registered via register_project. Attribution generated by generate_attribution.
Week 8
Launch week
Tally usage with the Dependency Graph. Feature the best entries on the Steam page and official channels. Scan for contamination with License Scanner.

What this campaign produces (example)

Tool

ToolIP Creator OS

docenty/ip-creator Skill

Agent Skills standard. Works across Claude Code, Codex, Cursor.

Install target 100

Mini-game30 vibe coders

30 Fan-made Mini-games

Built on the Phaser Ready-to-Build Pack. Auto-attributed.

Registered project target 30

Tool

ToolIP Creator OS

License Scanner Report

Detects unlicensed assets, missing attribution, outdated Packs; opens a fix PR.

Contamination incident target 0

Results dashboard (target mockup)

As of week 4

MCP calls4,500

Projects using assets30

30-day reuse rate40%

Missing attribution / unlicensed0

Wishlist traffic8,000

What we say in the meeting

“We measure MCP calls, projects using assets, and 30-day reuse rate, not website visitors. You can start today with a CC0 pack, no famous IP required.”

Proposed package

Free for developers · rights-holder SaaS subscription + transaction fee (hypothesis)

A concept scenario built from a fictional brand, buyer and schedule. Numbers are campaign targets, not achieved results or customer cases. Prices are a sales hypothesis (VAT excluded).

How a fan builds one

One line in, a submittable app 90 seconds later

Same in the web builder or in the Claude Code / Codex you already use. The agent fetches the IP Pack and rules; the fan writes one line of idea.

① Write one line
"Make a Crunch Lab member match test"
② The agent pulls the Pack and rules
Allowed assets · banned phrases · credits applied automatically
③ Preview → submit
Rule check shown as passed · one button into the approval queue

⏱ 90 s

fan-vibes.docenty.ai/build

IP PackCrunch Lab v2.1 · fan creation allowed

6 allowed · 4 banned · commerce: supporter pack only

Preview

Night Shift Researcher

Share

✓ Rules passed · credits added · ready to submit

✕ No account to create

Submit via link. Connect Docenty Pay only when payouts matter

✕ No design assets

The official Pack supplies characters, palette and prompts

✕ Less rule-checking

Allowed and banned come as code; the agent filters first (not a substitute for legal review)

Prompt → app → submit. Numbers are demo targets.

Developer GTM (concept)

npm for AI-ready IP

One install and your coding agent handles everything from rights checks to applying the content. The KPIs are MCP calls, projects using assets and 30-day reuse rate, not website visitors.

# From Claude Code / Codex / Cursor
$ npx skills add docenty/ip-creator          # SKILL.md + assets + LICENSE (Agent Skills standard)

# 5 MCP tools (v0)
search_assets(intent, usage, budget)         # Intent-based search: compare rights, format, price and engine compatibility at once
check_license(asset_id, project_context)     # License-as-Code verdict — anything outside the rules is "approval required"
get_asset_pack(asset_id, framework)          # Ready-to-Build Pack for Phaser / Godot / React
generate_attribution(asset_ids)              # Auto-generated attribution
register_project(asset_ids, project_url)     # Register in the Dependency Graph → tracking and payouts

Concept mockup · not a deliverable

Dependabot for IP

Checks each project for license changes, regional restrictions, missing attribution and new Pack versions, then opens fix PRs. Entry point: the free License Scanner.

Agent-native IP Request

When an agent can't find a fitting asset, it aggregates demand (type, use, medium, country, budget, deadline) with consent → a sales asset for suppliers.