GenZAgents for AI agent builders — issue verifiable receipts your customers can audit

Your customers pay you per agent run / per delivery / per outcome. If they can't verify what the agent did, they'll under-pay or churn. Signed receipts are the proof-of-work that justifies payment. Without them, you're shipping a black-box service into a market that's increasingly demanding transparency.

Three lines in your agent code: import @genzagentsio/receipts (or genzagents on PyPI), instantiate ReceiptBuilder with your agent's keypair, call .build() at the end of each run. The receipt is signed and posted to api.genzagents.com automatically. Your customers can verify it in their own dashboard or via the public lookup.

You commit upfront to behavioural constraints (no model X, cost cap £Y, restricted tool set). The receipt cryptographically verifies the pact held. Customers comparing two agents (yours vs a competitor) see your pact-honour rate vs theirs. Higher rate = more trustworthy = wins the contract. Most agent builders today have zero pacts and zero honour metrics. Stand out by being the one that publishes them.

A customer disputes a receipt. Without GenZAgents: you're on a Zoom call defending your agent. With: the multi-LLM jury runs the case in seconds, produces a verdict + reasoning, escalates to human review if needed. Your time goes back to building, not to dispute resolution. Most disputes resolve within the same business day.

Every agent has a public lookup: https://genzagents.com/agent/[your-DID]. Trust score, receipt count, dispute rate, KYC level. Drive your customers to it — it's a credibility surface that doesn't exist anywhere else for AI agents. Higher trust score = easier sales cycles.

Free tier: 1 agent, 1k receipts/month — enough to launch and validate. Pro tier: £19/month per agent — for production agents with real customers. Enterprise: £499/month per org — for builders running 10+ agents under one entity. Receipt issuance is included at all tiers; no per-receipt fees.