PRODUCTCLANK/ HOW IT WORKS
SPEC · the composable engine
network × action × reward → we execute
What ProductClank actually is

A composable distribution engine.

Any brand composes a campaign — pick a network, an action, and a reward — and we mobilize the community that executes it. The examples on this page are just presets: every campaign is custom, and you pay only for the outcome you care about.

And because it's callable from the web, a Farcaster mini-app, or your own agent (API · MCP · skill), distribution becomes something your software triggers — not a dashboard someone logs into.

where · networkwhat · actionhow · interfaceearn · reward railwho-pays · sponsor
§00

The invariant loop

This spine never changes — it's what stays constant while the five axes below vary. Every product surface is a way of instrumenting one step of it.

01 · fund
Sponsor funds
Credits, USDC, a stream, or an ecosystem grant.
02 · define
Builder scopes a campaign
Picks network × action + criteria.
03 · discover
Community finds targets
AI surfaces the posts / repos / channels worth acting on.
04 · act
Members take action
Reply, star, comment, create.
05 · verify
Proof is checked
Author-match, AI-vision, or on-chain.
06 · settle
Value flows back
Crypto or points, to the actor.
settled value + reputation feeds the next campaign
§01

The five axes

Orthogonal by design. A campaign is a coordinate: pick one (or a set) on each axis. Roadmap work is almost always 'add a value to one axis' — which is exactly what can be templatized (see §08).

Axis A · WhereNetworkthe distribution surface
XLinkedInInstagramYouTubeRedditGitHubTikTok · research
Axis B · WhatActionthe unit of engagement
Repost / shareReply / commentLike / upvoteStar repoCreate · freestyle
Axis C · HowInterfacehow humans & agents plug in
Web app · createMini-app · act & earnAgent skill + MCP · both
Axis D · EarnReward railhow value settles
0x Splits poolSuperfluid pointsCreditsProduct-Line MeshRevenue pools
Axis E · Who-paysSponsorthe funding model
Superfluid · generalBuilder USDC · per-campaignEcosystem · privateRev-share partners
campaign = network × action × interface × reward_rail × sponsor  +  criteria(keywords, targets, limits)
§02

Network × Action — the coverage matrix

Where axis A meets axis B. Hover a cell to read its coordinate. The shape of this grid is the roadmap: a new network is a column, a new action is a row, and most cells are the same three verification patterns underneath.

action ↓   network →XLinkedInInstagramYouTubeRedditGitHub
Repost / share● live● live○ next
Reply / comment● live● live● live● live● live○ next
Like / upvote● live● live● live○ next● live
Star repo● live
Create · freestyle● live● live● live● live● live○ next
coordinate:hover a cell to inspect — network × action → status
live · shipped ○ next · natural extension — not applicableconceptual model — exact live/beta status shifts per network
+ new network = a column
GitHub added a whole column but lit only one cell (Star). Adding a network is: a discovery source + an action verifier + one interface binding.
+ new action = a row
'Star repo' was a new row across the matrix. Create · freestyle is the row that fills everywhere — content is network-agnostic.
verification collapses to 3
Every cell resolves to author-match, AI-vision, or on-chain proof. That's why cells are cheap to add — the hard parts are already built.
§03

Example campaigns — these are just presets

Every card below is one coordinate through the axes. They aren't a fixed menu: describe the outcome you want and we compose the campaign for it — you pay for that outcome, nothing else.

X·reply·$PRO
Authentic buzz on your launch tweet
GitHub·star·$PRO
Push your repo up the charts
YouTube·comment·points
Seed real discussion on your video
Reddit·comment·points
Show up in the right subreddits
Any·create video·USDC pool
Original content made about you
X · LinkedIn·repost·credits
Amplify your announcement
compose your own
Describe an outcome in plain English → the Concierge assembles the campaign. Or POST it straight from your own agent.
§04

Discovery & research

Before anyone acts, the platform finds the surface worth acting on — the same 'add a network' pattern, one layer up: each network gets a way to find who and where.

R·01
X creators
Relevant accounts from lists + keywords.
live
R·02
Reddit subreddits
Relevant communities to engage in.
live
R·03
YouTube channels
Relevant channels & creators.
live
R·04
TikTok creators
Creator research — built, not yet lit.
staged
§05

Distribution as an API — not a dashboard

The same capabilities, exposed to three kinds of operator. This is the wedge: traditional growth apps are places a human logs into; here a campaign is something your own agent or workflow can trigger. Every new capability should reach all three surfaces.

C·01
Web app
Builders create & manage campaigns, buy credits, watch results. The control room.
create · manage · fund
C·02
Farcaster mini-app
Where the community takes action, claims, and earns crypto or points. The action floor.
act · verify · claim
C·03
Agent skill + MCP server
Autonomous agents create campaigns and participate programmatically — same rails, no browser.
create + participate · API
§06

Value settlement — the reward rails

Axis D, told as evolution. Each rail was added without throwing away the last — they compose. Credits sit on the builder's side; crypto and points settle on the earner's side.

01
On-chain reward pools · 0x Splits
The origin. A pool of crypto split across verified participants.
crypto · Base
02
Superfluid leaderboard
Earn points on a live leaderboard; claim the real reward through the Superfluid app as a stream.
points → stream
03
Credit system
Builders buy credits with Stripe or crypto and spend them to create campaigns; earners on the other side receive crypto or Superfluid points. The connective tissue between fiat-in and value-out.
Stripe / crypto in
04
Product-Line Mesh
A Supabase-built settlement fabric letting builders, agents & projects stream value to each other — any token, any time, any direction.
any token · any time
05
On-chain revenue pools future
Participants share in the proceeds of a campaign's success — ongoing upside, not a one-time reward. Rides the same infrastructure with its own distribution attributes.
rev-share
§07

Who funds the reward — the sponsor models

Axis E is the business model. Same actions, same rails; what changes is the source of the money and how tightly it's bound to a single campaign.

E·01 · general
Superfluid — platform sponsor
A standing reward source not attached to any one campaign. Keeps the flywheel spinning by default.
live · platform-wide
E·02 · per-campaign
Builder-funded USDC
A builder deposits USDC into one specific campaign and distributes it to participants. Live in the mini-app; web app support pending.
mini-app · web pending
E·03 · private
Ecosystem / private campaigns
An ecosystem funds a bounded growth push for its own participants — e.g. GoodDollar's two-week campaign.
live · e.g. GoodDollar
E·04 · partnership
Revenue-share partners future
Ongoing, tokenized revenue sharing instead of one-time rewards — the CAC direction.
future · ongoing upside
§08

The repeated shapes → templatable skills

This is the payoff of drawing it this way. Because the axes are orthogonal, most future work is one of five moves — and each move is the same checklist every time. That's a skill.

skill · add-network
Add a network
Discovery source + action verifier(s) + one interface binding + credit cost. Fills a column.
column · axis A
skill · add-action
Add an action type
Claim flow + one of 3 verification methods + reward mapping. Fills a row.
row · axis B
skill · bind-interface
Expose to a surface
Take an existing capability and wire it into web / mini-app / MCP so all three stay in parity.
binding · axis C
skill · add-rail
Add a reward rail
A settlement adapter — Splits / Superfluid / credits / Mesh — behind the same 'value settles' step.
adapter · axis D
skill · add-sponsor
Add a sponsor model
A funding source + a distribution policy. General, per-campaign, private, or rev-share.
policy · axis E
the thesis
Configuration, not rewrites
If a request maps to one of the five moves above, it's a template — not a project. That's the platform's leverage.
generalize everything
§09

How it grew

Read left to right, each stage is 'we generalized one axis.' The story is not more features — it's the same loop, widened.

startX only — repost & reply
widen AMulti-network: LinkedIn, IG, YouTube, Reddit, GitHub
widen BMulti-action: comment, star, freestyle create
+ discoveryCreator / subreddit / channel research
widen CMini-app to act, then agent skill + MCP
widen DSplits → Superfluid → credits → Mesh
widen EGeneral, builder-USDC & private sponsors
nextOn-chain revenue pools · shared upside
§10

How to describe it — three altitudes

Pull the level to match the room. Same platform, focused for the listener.

One line
ProductClank turns communities into a distribution engine — and pays them for the results.
Elevator
Pick a network, an action, and a reward; we mobilize the community that executes — replies, comments, stars, videos — and settle value on-chain. Every campaign is custom, and you pay for the outcome you care about.
Technical
A composable campaign primitive — network × action × interface × reward-rail × sponsor — callable from the web, a Farcaster mini-app, or your own agent. Distribution as an API, not a dashboard.