Agent capabilities you can inspect before you buy

CrossingKey Intelligence provides defined paid utilities for AI agents through its public MCP. Every capability should make the task, input, output, price, requirements, approval rule, and verification path visible before use.

MCP endpoint: https://mcp.crossingkeyintelligence.com/mcp
Machine-readable discovery: mcp.json

Available capabilities

Artifact Integrity Manifest — $0.10 USDC

Purpose: Create deterministic SHA-256 entries and a root manifest hash for supplied text artifacts.

Input: Text artifacts and identifiers.
Output: Per-artifact hashes, manifest hash, and verification metadata.
Requirements: UTF-8 text and stable artifact names.
Approval: Human approval is required before payment.
Example result: A manifest containing each artifact’s SHA-256 hash and one reproducible root hash.

MCP Schema Audit — $1.00 USDC

Purpose: Review MCP tool schemas for clear descriptions and bounded input contracts.

Input: MCP endpoint or supplied schema data.
Output: Findings grouped by severity with affected tools and recommended corrections.
Requirements: Publicly reachable MCP surface or valid schema payload.
Approval: Human approval is required before payment.
Example result: A finding that a tool lacks a required input description, with the exact field and remediation.

x402 Compatibility Audit — $1.00 USDC

Purpose: Check a public x402 endpoint for payment protocol compatibility defects.

Input: Public endpoint URL.
Output: Compatibility findings, observed headers, and protocol checks.
Requirements: HTTPS endpoint accessible to the audit service.
Approval: Human approval is required before payment.
Example result: A report showing whether the endpoint advertises and responds to the expected x402 flow.

OpenAPI Quality Audit — $1.00 USDC

Purpose: Evaluate an OpenAPI document’s structure and required fields.

Input: OpenAPI JSON or YAML document.
Output: Validation findings with affected paths and fields.
Requirements: OpenAPI document supplied in a supported format.
Approval: Human approval is required before payment.
Example result: A list of missing operation descriptions or malformed request schemas.

Machine-Commerce Readiness Audit — $2.00 USDC

Purpose: Check a public machine-commerce service URL for readiness indicators.

Input: Public service URL.
Output: Readiness findings and observed protocol signals.
Requirements: HTTPS endpoint and publicly observable service metadata.
Approval: Human approval is required before payment.
Example result: A readiness report showing which discovery, payment, receipt, and verification signals were observed.

How access works

1. Inspect the machine-readable capability listing.
2. Choose the capability and review its requirements.
3. Provide the required input.
4. Review the payment challenge before authorizing anything.
5. Approve payment through a wallet or client you control.
6. Review the result and verify it against your requirements.

Do not give an agent unrestricted spending authority. CrossingKey’s wallet is configured to receive payments only. A capability result is not a security certification, legal opinion, or guarantee of business outcomes.

Need a tool for yourself?

Browse the ready-to-use guides, checklists, and script library. Those products use Shopify pricing and delivery terms separately from the USDC-priced MCP utilities.