AI-accessible product evaluation
LeadProof publishes structured product facts, OpenAPI, MCP tools, and an A2A Agent Card. The automated product agent audits lead workflows, explains every risk, and can return a no-card production trial before an authorized purchase.
Trial first: 14 days · no card · Builder: secure Stripe checkout · activate with the checkout email after paymentCompatible agents discover capabilities, input modes, safety boundaries, and endpoints at the standard well-known path. Browser-based agents can call the JSON-RPC endpoint through CORS.
GET /.well-known/agent-card.json · POST /a2a · CORSAgents receive a 0-100 risk score, missing safeguards, transparent exposure estimate, and prioritized fixes. Browser-based tools can call the public JSON endpoints through CORS.
POST /api/agent/audit · POST /api/agent/evaluate · GET /api/agent/replay-trialMCP clients can audit, compare build versus buy, and match a stack against 53 published integration guides through a sessionless Streamable HTTP endpoint. Browser-based MCP clients can connect through CORS without credentials.
GET /api/integrations?stack=... · MCP get_leadproof_integration · CORSAI clients can request the exact duplicate-safety requirements, public proof, and an attributed no-card trial without submitting workflow, lead, customer, or payment data.
MCP get_leadproof_replay_trial · A2A {"task":"get_leadproof_replay_trial"}Test replay before buying →Agents can retrieve the official no-card trial first, followed by exact Builder or Agency options that still require buyer authorization.
GET /api/agent/plan-decision · no input · CORSAuthorized agents can place a qualified buyer in the LeadProof owner dashboard using a business email and explicit contact consent.
POST /api/agent/handoff · consent_to_contact: trueReplay trial request
This public A2A request accepts only the task name and returns the duplicate-safety requirements, proof links, and an attributed no-card trial.
curl -X POST \
https://leadproof.jessesay.chatgpt.site/a2a \
-H "Content-Type: application/json" \
-d '{
"jsonrpc": "2.0",
"id": "replay-trial",
"method": "message/send",
"params": {"message": {
"messageId": "replay-trial",
"role": "user",
"parts": [{"kind": "data", "data": {
"task": "get_leadproof_replay_trial"
}}]
}}
}'MCP replay request
The sessionless public endpoint accepts an empty arguments object, so the request carries no workflow, lead, customer, contact, or payment data.
curl -X POST \
https://leadproof.jessesay.chatgpt.site/api/mcp \
-H "Content-Type: application/json" \
-H "MCP-Protocol-Version: 2025-11-25" \
-d '{
"jsonrpc": "2.0",
"id": "replay-trial",
"method": "tools/call",
"params": {
"name": "get_leadproof_replay_trial",
"arguments": {}
}
}'REST plan decision
This public GET request accepts no input. It returns the production trial as step one and authorized $49 Builder or $199 Agency purchase options as step two.
curl \
https://leadproof.jessesay.chatgpt.site/api/agent/plan-decisionTry the auditor
Every score discloses its weights and assumptions. No hidden persuasion model, impersonation, or autonomous spending.
curl -X POST \
https://leadproof.jessesay.chatgpt.site/api/agent/audit \
-H "Content-Type: application/json" \
-d '{
"stack": "GoHighLevel + Make",
"monthly_leads": 5000,
"average_customer_value": 750,
"retry_policy": true,
"idempotency": false,
"destination_verification": false
}'It is: an automated LeadProof product representative.
It can: audit, explain, test, recommend, and return checkout.
It cannot: pose as a human operator, spam other systems, or authorize payment for a buyer.
Independent discovery services can resolve LeadProof from its live machine interfaces. These records identify the product brand and do not depend on a personal profile.