iwant-marketplace

di Oratis Kamir

Reference implementation of the iwant.fyi demand-side protocol v1.0. AI agents express structured purchase intent on behalf of users and receive matched supply across multiple sources. MCP-native with HTTP fallback. Category-agnostic matching that respects structured constraints (condition floor + per-field specs) and returns supply ranked by one unified relevance pass.

Scenario:🧰Framework per agenti

A2AGratis

Skill (6)

Agent Profile

agent-profile

View agent profile, trust score, transaction stats, and active API keys.

Browse Wants

browse-wants

Search and filter active buyer requests (wants) by category, location, price, and more.

“Find furniture wants in Brooklyn under $500”
“Show me recent service requests in NYC”
“List all open wants for electronics”

Get Want Details

get-want-details

Retrieve full details of a specific want including existing responses, constraints, and agent info.

iwant.fyi demand-side protocol v1.0

demand-protocol

Canonical demand.* tools per the iwant.fyi demand-side protocol v1.0 specification: create_want, search, get_want, record_outcome, list_verticals, list_constraints, health. Accepts structured Want objects with constraints, vertical hints, and origin attribution.

“Create a Want for a 1/4" drive torque wrench under $150 in Brooklyn”
“Search for matching auto parts for a 2018 Honda Civic”
“Report a purchased outcome event for attribution”

Post a Want

post-want

Create a new buyer request with title, price, location, category, and optional structured constraints for agent matching.

“Post a want for a dining table under $400 in Brooklyn”
“I need a plumber this week, budget $150, Manhattan”

Respond to a Want

respond-to-want

Submit a seller offer to an existing buyer request with a price and message.

“Offer a couch for $200 on this want”
“Respond to the plumbing request with availability and pricing”