I Built 300 Utility APIs for AI Agents with Cloudflare Workers and x402

#api #webdev #ai #showdev

What started as a small collection of developer utilities has grown into Koyori Utils API — a service offering 300 API endpoints.

The goal is simple: give developers and AI agents access to practical data-processing tools through a single service, without requiring a monthly subscription.

What does it offer?

The collection covers several common development tasks:

JSON utilities: formatting, validation, schema-related tools, and data transformation
CSV processing: converting and analyzing structured data
Text utilities: extraction, cleanup, comparison, and formatting
Array operations: manipulating and transforming collections
Math and statistics: calculations and numerical analysis
Date and time: parsing, conversion, and date-related operations

Instead of maintaining separate small utilities for every workflow, an AI agent can discover available operations through the service's OpenAPI specification.

The technology

Koyori Utils API is built using:

Cloudflare Workers for serverless deployment
TypeScript and Hono for API development
x402 on Stacks mainnet for per-request payments
OpenAPI for machine-readable documentation

The current listed price is 0.001 STX per paid request, with no monthly subscription required.

How do requests work?

For example, the service includes a SHA-256 hashing endpoint:

curl -X POST \
  https://koyori-utils-api.koyori-aibtc.workers.dev/api/hash \
  -H "Content-Type: application/json" \
  -d '{"text":"hello"}'

An unauthenticated request returns an HTTP 402 Payment Required response. A compatible client needs to complete the x402 payment flow to access the paid result.

From 72 endpoints to 300

One of the biggest lessons from expanding the project was the importance of automation.

Rather than adding every utility manually through repeated deployment steps, the development workflow now includes:

Creating groups of new utility endpoints
Checking for duplicate API routes
Updating homepage listings and documentation
Running TypeScript validation
Creating backups before local updates
Deploying and checking the live service

The production homepage now lists 300 endpoints, and the service health check is responding normally.

There is still important work ahead, especially broader endpoint testing, payment compatibility verification, and gathering real user feedback.

Try exploring the project

Website: https://koyori-utils-api.koyori-aibtc.workers.dev

OpenAPI specification: https://koyori-utils-api.koyori-aibtc.workers.dev/openapi.json

Source code: https://github.com/sophiaalice0913-cmyk/koyori-utils-api

I'd especially appreciate feedback on API discoverability, developer experience, and the kinds of utilities AI agents actually need.

If you were building an autonomous agent, which utility would you want available as a simple API call?

Thanks for reading! 💗

DEV Community