Scope and output
Know exactly what is being reviewed.
One public agent-facing service, API, MCP server, or paid endpoint.
- Agent discovery and crawlability review
- llms.txt, schema, OpenAPI, MCP, and capability-contract checks
- Offer, checkout, x402, and payment-proof review where present
- Buyer-journey test from first request through fulfillment
- Trust, provenance, privacy, and failure-mode findings
- Prioritized agent-readiness remediation plan
- Price
- $199 USDC or the signed exact-ETH alternative
- Delivery target
- 3 business days after confirmed payment
- Boundary
- Testing is passive and non-destructive. Paid third-party requests, credentials, private endpoints, and production mutations are excluded unless separately authorized and funded.
Decision questions
The report is built to answer these.
- Can an unfamiliar agent discover and understand the capability?
- Are prices, permissions, proof types, and failure states machine-readable?
- Can payment intent be bound safely to fulfillment?
- Does the complete buyer journey fail closed without stranding the buyer?
Good fit
Who should buy this review?
- MCP server operators adding paid capabilities
- API businesses designing agent-facing purchase flows
- x402 or wallet-checkout teams preparing for external buyers
- Agent marketplaces that need inspectable trust and fulfillment contracts
Direct answers
Questions before ordering
What is an agent-commerce audit?
It tests the public journey an AI buyer must complete: discovery, capability understanding, trust evaluation, offer retrieval, checkout, payment-proof handling, fulfillment, and recovery from expected failures.
Which systems can be reviewed?
The standard scope covers one public MCP server, API, agent-facing service, or paid endpoint. It can evaluate llms.txt, OpenAPI, MCP metadata, structured data, x402-style contracts, and visible payment flows where present.
Will the audit spend money or mutate production?
No. Standard testing is passive and non-destructive. Paid third-party requests, credentials, private endpoints, and production mutations require separate authorization and funding.
Does KnownFix support both Base Pay and direct transactions?
KnownFix can review both patterns. It treats Base Pay UserOperation hashes and direct transaction hashes as different evidence types and checks whether intent, amount, recipient, chain, and replay controls remain bound.
What is delivered?
The $199 engagement returns buyer-journey evidence, prioritized trust and failure-mode findings, and a remediation plan in portable HTML and verified Markdown.
Compare scope
Other KnownFix reviews
Start with a bounded question.
Choose this review only when its published scope matches the decision you need to make. The order target and notes remain private unless you explicitly publish the target.